Version:	3.1.18
Date:	2025-10-19
Title:	Functional Analysis of Phenotypic Growth Data to Smooth and Extract Traits
Depends:	R (≥ 3.5.0)
Imports:	dae, dplyr, GGally, ggplot2, grDevices, Hmisc, JOPS, methods, RColorBrewer, readxl, reshape, stats, stringi, utils
Suggests:	testthat, nlme, R.rsp, scales
VignetteBuilder:	R.rsp
Description:	Assists in the plotting and functional smoothing of traits measured over time and the extraction of features from these traits, implementing the SET (Smoothing and Extraction of Traits) method described in Brien et al. (2020) Plant Methods, 16. Smoothing of growth trends for individual plants using natural cubic smoothing splines or P-splines is available for removing transient effects and segmented smoothing is available to deal with discontinuities in growth trends. There are graphical tools for assessing the adequacy of trait smoothing, both when using this and other packages, such as those that fit nonlinear growth models. A range of per-unit (plant, pot, plot) growth traits or features can be extracted from the data, including single time points, interval growth rates and other growth statistics, such as maximum growth or days to maximum growth. The package also has tools adapted to inputting data from high-throughput phenotyping facilities, such from a Lemna-Tec Scananalyzer 3D (see https://www.youtube.com/watch?v=MRAF_mAEa7E/ for more information). The package 'growthPheno' can also be installed from http://chris.brien.name/rpackages/.
License:	GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
URL:	http://chris.brien.name/
BugReports:	https://github.com/briencj/growthPheno/issues
RoxygenNote:	5.0.1
NeedsCompilation:	no
Packaged:	2025-10-19 01:07:52 UTC; briencj
Author:	Chris Brien [aut, cre]
Maintainer:	Chris Brien <chris.brien@adelaide.edu.au>
Repository:	CRAN
Date/Publication:	2025-10-19 05:30:02 UTC

Functional Analysis of Phenotypic Growth Data to Smooth and Extract Traits

Description

Assists in the plotting and functional smoothing of traits measured over time and the extraction of features from these traits, implementing the SET (Smoothing and Extraction of Traits) method described in Brien et al. (2020) Plant Methods, 16. Smoothing of growth trends for individual plants using natural cubic smoothing splines or P-splines is available for removing transient effects and segmented smoothing is available to deal with discontinuities in growth trends. There are graphical tools for assessing the adequacy of trait smoothing, both when using this and other packages, such as those that fit nonlinear growth models. A range of per-unit (plant, pot, plot) growth traits or features can be extracted from the data, including single time points, interval growth rates and other growth statistics, such as maximum growth or days to maximum growth. The package also has tools adapted to inputting data from high-throughput phenotyping facilities, such from a Lemna-Tec Scananalyzer 3D (see <https://www.youtube.com/watch?v=MRAF_mAEa7E/> for more information). The package 'growthPheno' can also be installed from <http://chris.brien.name/rpackages/>.

Version: 3.1.18

Date: 2025-10-19

Index

The following list of functions does not include those that are soft-deprecated, i.e. those that have been available in previous versions of growthPheno but will be removed in future versions. For a description of the use of the listed functions and vignettes that are available, see the Overview section below.

Overview

This package can be used to perform a functional analysis of growth data using splines to smooth the trend of individual plant traces over time and then to extract features or tertiarty traits for further analysis. This process is called smoothing and extraction of traits (SET) by Brien et al. (2020), who detail the use of growthPheno for carrying out the method. However, growthPheno now has the two wrapper, or primary, functions traitSmooth and traitExtractFeatures that implement the SET approach. These may be the only functions that are used in that the complete SET process can be carried out using only them. The Tomato vignette illustrates their use for the example presented in Brien et al. (2020).

The function traitSmooth utilizes the secondary functions probeSmooths, plotSmoothsComparison and plotSmoothsMedianDevns and accepts the arguments of the secondary functions. The function probeSmooths utilizes the tertiary functions byIndv4Times_SplinesGRs and byIndv4Times_GRsDiff, which in turn call the function smoothSpline. The function plotSmoothsComparison calls plotDeviationsBoxes. All of these functions play a role in choosing the smoothing method and parameters for a data set.

The primary function traitExtractFeatures uses the secondary functions getTimesSubset and the set of byIndv4Intvl_ functions. These functions are concerned with the extraction of traits that yield a single value for each individual in the data.

Recourse to the secondary and terriary functions may be necessary for special cases. Their use is illustrated in the Rice vignette.

Use vignette("Tomato", package = "growthPheno") or vignette("Rice", package = "growthPheno") to access either of the vignettes.

In addition to functions that implement SET approach, growthPheno also has functions for importing and organizing the data that are generally applicable, although they do have defaults that make them particularly adapted to data from a high-throughput phenotyping facility based on a Lemna-Tec Scananalyzer 3D system.

Data suitable for use with this package consists of columns of data obtained from a set of individuals (e.g. plants, pots, carts, plots or units) over time. There should be a unique identifier for each individual and a time variable, such as Days after Planting (DAP), that contain no repeats for an individual. The combination of the identifier and a time for an individual should be unique to that individual. For imaging data, the individuals may be arranged in a grid of Lanes \times Positions. That is, the minimum set of columns is an individuals, a times and one or more primary trait columns.

Author(s)

Chris Brien [aut, cre] (<https://orcid.org/0000-0003-0581-1817>)

Maintainer: Chris Brien <chris.brien@adelaide.edu.au>

References

Brien, C., Jewell, N., Garnett, T., Watts-Williams, S. J., & Berger, B. (2020). Smoothing and extraction of traits in the growth analysis of noninvasive phenotypic data. Plant Methods, 16, 36. doi:10.1186/s13007-020-00577-6.

See Also

dae

Calculates growth rates (AGR, PGR, RGRdiff) between pairs of values in a vector

Description

Calculates either the Absolute Growth Rate (AGR), Proportionate Growth Rate (PGR) or Relative Growth Rate (RGR) between pairs of time points, the second of which is lag positions before the first. in x.

Usage

AGRdiff(x, time.diffs, lag=1)
PGR(x, time.diffs, lag=1)
RGRdiff(x, time.diffs, lag=1)

Arguments

Details

The AGRdiff is calculated as the difference between a pair of values divided by the time.diffs. The PGR is calculated as the ratio of a value to a second value which is lag values ahead of the first in x and the ratio raised to the power of the reciprocal of time.diffs. The RGRdiff is calculated as the log of the PGR and so is equal to the difference between the logarithms of a pair of values divided by the time.diffs. The differences and ratios are obtained using calcLagged with lag = 1.

Value

A numeric containing the growth rates which is the same length as x and in which the first lag values NA.

Author(s)

Chris Brien

See Also

byIndv4Intvl_GRsAvg, byIndv4Intvl_GRsDiff, byIndv4Times_GRsDiff, byIndv4Times_SplinesGRs,
calcLagged

Examples

data(exampleData)
longi.dat$PSA.AGR <- with(longi.dat, AGRdiff(PSA, time.diffs = DAP.diffs))

Selects a subset of variables using Principal Variable Analysis (PVA)

Description

Principal Variable Analysis (PVA) (Cumming and Wooff, 2007) selects a subset from a set of the variables such that the variables in the subset are as uncorrelated as possible, in an effort to ensure that all aspects of the variation in the data are covered.

Usage

PVA(obj, ...)

Arguments

Details

PVA is the generic function for the PVA method. Use methods("PVA") to get all the methods for the PVA generic.

PVA.data.frame is a method for a data.frame.

PVA.matrix is a method for a matrix.

Value

A data.frame giving the results of the variable selection. It will contain the columns Variable, Selected, h.partial, Added.Propn and Cumulative.Propn.

Author(s)

Chris Brien

References

Cumming, J. A. and D. A. Wooff (2007) Dimension reduction via principal variables. Computational Statistics and Data Analysis, 52, 550–565.

See Also

PVA.data.frame, PVA.matrix, intervalPVA, rcontrib

Selects a subset of variables stored in a data.frame using Principal Variable Analysis (PVA)

Description

Principal Variable Analysis (PVA) (Cumming and Wooff, 2007) selects a subset from a set of the variables such that the variables in the subset are as uncorrelated as possible, in an effort to ensure that all aspects of the variation in the data are covered.

Usage

## S3 method for class 'data.frame'
PVA(obj, responses, nvarselect = NULL, p.variance = 1, include = NULL, 
    plot = TRUE, ...)

Arguments

Details

The variable that is most correlated with the other variables is selected first for inclusion. The partial correlation for each of the remaining variables, given the first selected variable, is calculated and the most correlated of these variables is selects for inclusion next. Then the partial correlations are adjust for the second included variables. This process is repeated until the specified criteria have been satisfied. The possibilities are:

1. the default (nvarselect = NULL and p.variance = 1), which selects all variables in increasing order of amount of information they provide;

2. to select exactly nvarselect variables;

3. to select just enough variables, up to a maximum of nvarselect variables, to explain at least p.variance*100 per cent of the total variance.

Value

A data.frame giving the results of the variable selection. It will contain the columns Variable, Selected, h.partial, Added.Propn and Cumulative.Propn.

Author(s)

Chris Brien

References

Cumming, J. A. and D. A. Wooff (2007) Dimension reduction via principal variables. Computational Statistics and Data Analysis, 52, 550–565.

See Also

PVA, PVA.matrix, intervalPVA.data.frame, rcontrib

Examples

data(exampleData)
longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1)
longi.dat <- within(longi.dat, 
                    {
                      Max.Height <- pmax(Max.Dist.Above.Horizon.Line.SV1,  
                                         Max.Dist.Above.Horizon.Line.SV2)
                      Density <- PSA/Max.Height
                      PSA.SV = (PSA.SV1 + PSA.SV2) / 2
                      Image.Biomass = PSA.SV * (PSA.TV^0.5)
                      Centre.Mass <- (Center.Of.Mass.Y.SV1 + Center.Of.Mass.Y.SV2) / 2
                      Compactness.SV = (Compactness.SV1 + Compactness.SV2) / 2
                    })
responses <- c("PSA","PSA.SV","PSA.TV", "Image.Biomass", "Max.Height","Centre.Mass",
               "Density", "Compactness.TV", "Compactness.SV")
results <-  PVA(longi.dat, responses, p.variance=0.9, plot = FALSE)

Selects a subset of variables using Principal Variable Analysis (PVA) based on a correlation matrix

Description

Principal Variable Analysis (PVA) (Cumming and Wooff, 2007) selects a subset from a set of the variables such that the variables in the subset are as uncorrelated as possible, in an effort to ensure that all aspects of the variation in the data are covered.

Usage

## S3 method for class 'matrix'
PVA(obj, responses, nvarselect = NULL, p.variance = 1, include = NULL, 
    plot = TRUE, ...)

Arguments

Details

The variable that is most correlated with the other variables is selected first for inclusion. The partial correlation for each of the remaining variables, given the first selected variable, is calculated and the most correlated of these variables is selects for inclusion next. Then the partial correlations are adjust for the second included variables. This process is repeated until the specified criteria have been satisfied. The possibilities are:

1. the default (nvarselect = NULL and p.variance = 1), which selects all variables in increasing order of amount of information they provide;

2. to select exactly nvarselect variables;

3. to select just enough variables, up to a maximum of nvarselect variables, to explain at least p.variance*100 per cent of the total variance.

Value

A data.frame giving the results of the variable selection. It will contain the columns Variable, Selected, h.partial, Added.Propn and Cumulative.Propn.

Author(s)

Chris Brien

References

Cumming, J. A. and D. A. Wooff (2007) Dimension reduction via principal variables. Computational Statistics and Data Analysis, 52, 550–565.

See Also

PVA, PVA.data.frame, intervalPVA.data.frame, rcontrib

Examples

data(exampleData)
longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1)
longi.dat <- within(longi.dat, 
                    {
                      Max.Height <- pmax(Max.Dist.Above.Horizon.Line.SV1,  
                                         Max.Dist.Above.Horizon.Line.SV2)
                      Density <- PSA/Max.Height
                      PSA.SV = (PSA.SV1 + PSA.SV2) / 2
                      Image.Biomass = PSA.SV * (PSA.TV^0.5)
                      Centre.Mass <- (Center.Of.Mass.Y.SV1 + Center.Of.Mass.Y.SV2) / 2
                      Compactness.SV = (Compactness.SV1 + Compactness.SV2) / 2
                    })
responses <- c("PSA","PSA.SV","PSA.TV", "Image.Biomass", "Max.Height","Centre.Mass",
               "Density", "Compactness.TV", "Compactness.SV")
R <- Hmisc::rcorr(as.matrix(longi.dat[responses]))$r
results <-  PVA(R, responses, p.variance=0.9, plot = FALSE)

Prepped data from an experiment to investigate a rice germplasm panel.

Description

The data is the full set of Lanes and Positions from an experiment in a Smarthouse at the Plant Accelerator in Adelaide. It is used in the growthPheno-package as an executable example to illustrate the use of growthPheno. The experiment and data collection are described in Al-Tamimi et al. (2016) and the data is derived from the data.frame in the file 00-raw.0254.dat.rda that is available from Al-Tamimi et al. (2017); halpf od the unprred data is in RiceRaw.dat.

Usage

data(RicePrepped.dat)

Format

A data.frame containing 14784 observations on 32 variables. The names of the columns in the data.frame are:

Source

Al-Tamimi N, Brien C, Oakey H, Berger B, Saade S, Ho YS, Schmockel SM, Tester M, Negrao S. (2017) Data from: Salinity tolerance loci revealed in rice using high-throughput non-invasive phenotyping. Retrieved from: doi:10.5061/dryad.3118j.

References

Al-Tamimi, N, Brien, C.J., Oakey, H., Berger, B., Saade, S., Ho, Y. S., Schmockel, S. M., Tester, M. and Negrao, S. (2016) New salinity tolerance loci revealed in rice using high-throughput non-invasive phenotyping. Nature Communications, 7, 13342. Retrieved from doi:10.1038/ncomms13342.

Data for an experiment to investigate a rice germplasm panel

Description

The data is half (the first 12 of 24 Lanes) of that from an experiment in a Smarthouse at the Plant Accelerator in Adelaide. It is used in the growthPheno-package as an executable example to illustrate the use of growthPheno. The experiment and data collection are described in Al-Tamimi et al. (2016) and the data is derived from the data.frame in the file 00-raw.0255.dat.rda that is available from Al-Tamimi et al. (2017).

Usage

data(RiceRaw.dat)

Format

A data.frame containing 7392 observations on 34 variables.

Source

Al-Tamimi N, Brien C, Oakey H, Berger B, Saade S, Ho YS, Schmockel SM, Tester M, Negrao S: Data from: Salinity tolerance loci revealed in rice using high-throughput non-invasive phenotyping. Retrieved from: doi:10.5061/dryad.3118j.

References

Al-Tamimi, N, Brien, C.J., Oakey, H., Berger, B., Saade, S., Ho, Y. S., Schmockel, S. M., Tester, M. and Negrao, S. (2016) New salinity tolerance loci revealed in rice using high-throughput non-invasive phenotyping. Nature Communications, 7, 13342. Retrieved from doi:10.1038/ncomms13342.

Calculates the Water Use Index (WUI)

Description

Calculates the Water Use Index, returning NA if the water use is zero.

Usage

WUI(response, water)

Arguments

Value

A numeric containing the response divided by the water, unless water is zero in which case NA is returned.

Author(s)

Chris Brien

Examples

data(exampleData)
PSA.WUE <- with(longi.dat, WUI(PSA.AGR, WU))

Tests if any values in a vector are anomalous in being outside specified limits

Description

Test whether any values in x are less than the value of lower, if it is not NULL, or are greater than the value of upper, if it is not NULL, or both.

Usage

anom(x, lower=NULL, upper=NULL, na.rm = TRUE)

Arguments

Value

A logical indicating whether any values have been found to be outside the limits specified by lower or upper or both.

Author(s)

Chris Brien

Examples

data(exampleData)
anom.val <- anom(longi.dat$sPSA.AGR, lower=2.5)

Creates a list of the values for the options of profile plots for the chosen smooth

Description

Creates a list of the values for the options of profile plots (and boxplots facets) for comparing smooths. Note that plots.by, facet.x, facet.y and include.raw jointly define the organization of the plots. The default settings are optimized for traitSmooth.

Usage

args4chosen_plot(plots.by = NULL, 
                 facet.x = ".", facet.y = ".", 
                 include.raw = "no", 
                 collapse.facets.x = FALSE, collapse.facets.y = FALSE, 
                 facet.labeller = NULL, facet.scales = "fixed", 
                 breaks.spacing.x = -2, angle.x = 0, 
                 colour = "black", colour.column = NULL, 
                 colour.values = NULL, alpha = 0.3, 
                 addMediansWhiskers = TRUE,
                 ggplotFuncs = NULL, 
                 ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, plotSmoothsComparison and args4profile_plot.

Examples

args4chosen_plot(plots.by = "Type", 
                 facet.x = "Tuning", facet.y = c("Smarthouse", "Treatment.1"), 
                 include.raw = "facet.x", 
                 alpha  = 0.4, 
                 colour.column = "Method", 
                 colour.values = c("orange", "olivedrab"))

Creates a list of the values for the smoothing parameters for which a smooth is to be extracted

Description

Creates a list of the values for the smoothing parameters for which a single smooth is to be extracted. The default settings for these are optimized for traitSmooth.

Usage

args4chosen_smooth(smoothing.methods = "logarithmic", 
                   spline.types = "PS", 
                   df = NULL, 
                   lambdas = NULL,
                   combinations = "single", 
                   ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth and probeSmooths.

Examples

args4chosen_smooth(smoothing.methods = "direct", 
                   spline.types = "NCSS", df = 4)
args4chosen_smooth(smoothing.methods = "log", 
                   spline.types = "PS", lambdas = 0.36)

Creates a list of the values for the options of profile plots for comparing smooths

Description

Creates a list of the values for the options of deviations boxplots for comparing smooths. Note that plots.by, facet.x and facet.y jointly define the organization of the plots. The default settings are optimized for traitSmooth so that, if you want to change any of these from their default settings when using args4devnboxes_plot with a function other than traitSmooth, then it is recommended that you specify all of them to ensure that the complete set has been correctly specified. Otherwise, the default settings will be those shown here and these may be different to the default settings shown for the function with which you are using args4devnboxes_plot.

Usage

args4devnboxes_plot(plots.by = "Type", 
                    facet.x = c("Method", "Tuning"), facet.y = ".", 
                    collapse.facets.x = TRUE, collapse.facets.y = FALSE, 
                    facet.labeller = NULL, facet.scales = "fixed", 
                    angle.x = 0, 
                    which.plots = "none", 
                    ggplotFuncs = NULL, 
                    ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, plotSmoothsComparison and args4chosen_plot.

Examples

args4devnboxes_plot(plots.by = "Type", 
                    facet.x = "Tuning", 
                    facet.y = c("Smarthouse", "Treatment.1"), 
                    which.plots = "absolute")

Creates a list of the values for the options of median deviations plots for smooths

Description

Creates a list of the values for the options of median deviations plots for smooths. Note that the arguments plots.by, plots.group, facet.x and facet.y jointly define the organization of the plots. The default settings are optimized for traitSmooth so that, if you want to change any of these from their default settings when using args4meddevn_plot with a function other than traitSmooth, then it is recommended that you specify all of them to ensure that the complete set has been correctly specified. Otherwise, the default settings will be those shown here and these may be different to the default settings shown for the function with which you are using args4meddevn_plot.

Usage

args4meddevn_plot(plots.by = NULL, plots.group = "Tuning", 
                  facet.x = c("Method","Type"), facet.y = ".",  
                  facet.labeller = NULL, facet.scales = "free_x",
                  breaks.spacing.x = -4, angle.x = 0, 
                  colour.values = NULL, shape.values = NULL, 
                  alpha = 0.5, 
                  propn.note = TRUE, propn.types = NULL, 
                  ggplotFuncs = NULL, 
                  ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths and plotSmoothsMedianDevns.

Examples

args4meddevn_plot(plots.by = "Type", plots.group = "Tuning",  
                  facet.x = "Method", facet.y = ".",
                  propn.types = c(0.02,0.1, 0.2))

Creates a list of the values for the options of profile plots for comparing smooths

Description

Creates a list of the values for the options of profile plots for comparing smooths. Note that plots.by, facet.x, facet.y and include.raw jointly define the organization of the plots. The default settings are optimized for traitSmooth so that, if you want to change any of these from their default settings when using args4profile_plot with a function other than traitSmooth, then it is recommended that you specify all of them to ensure that the complete set has been correctly specified. Otherwise, the default settings will be those shown here and these may be different to the default settings shown for the function with which you are using args4profile_plot.

Usage

args4profile_plot(plots.by = "Type", 
                  facet.x = c("Method", "Tuning"), facet.y = ".", 
                  include.raw = "facet.x", 
                  collapse.facets.x = TRUE, collapse.facets.y = FALSE, 
                  facet.labeller = NULL, facet.scales = "fixed", 
                  breaks.spacing.x = -4, angle.x = 0, 
                  colour = "black", colour.column = NULL, 
                  colour.values = NULL, alpha = 0.3, 
                  addMediansWhiskers = TRUE,
                  ggplotFuncs = NULL, 
                  ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, plotSmoothsComparison and args4chosen_plot.

Examples

args4profile_plot(plots.by = "Type", 
                  facet.x = "Tuning", facet.y = c("Smarthouse", "Treatment.1"), 
                  include.raw = "facet.x", 
                  alpha  = 0.4, 
                  colour.column = "Method", 
                  colour.values = c("orange", "olivedrab"))

Creates a list of the values for the smoothing parameters to be passed to a smoothing function

Description

Creates a list of the values for the smoothing parameters to be passed to a smoothing function. Note that smoothing.methods, spline.types, df and lambdas are combined to define the set of smooths. The default settings are optimized for traitSmooth so that, if you want to change any of these from their default settings when using args4smoothing with a function other than traitSmooth, then it is recommended that you specify all of them to ensure that the complete set has been correctly specified. Otherwise, the default settings will be those shown here and these may be different to the default settings shown for the function with which you are using args4smoothing.

Usage

 args4smoothing(smoothing.methods = "logarithmic", 
                spline.types = c("NCSS","PS"), 
                df = 5:7, 
                lambdas = list(PS = round(10^c(-0.5, 0, 0.5, 1), 
                                          digits = 3)),
                smoothing.segments = NULL, 
                npspline.segments = NULL, 
                na.x.action="exclude", na.y.action = "trimx", 
                external.smooths = NULL, 
                correctBoundaries = FALSE, 
                combinations = "allvalid", 
                ...)

Arguments

Value

A named list.

Author(s)

Chris Brien

See Also

traitSmooth and probeSmooths.

Examples

args4smoothing(smoothing.methods = "direct", 
               spline.types = "NCSS", df = NULL, lambdas = NULL, 
               smoothing.segments = NULL, npspline.segments = NULL, 
               combinations = "allvalid")
args4smoothing(smoothing.methods = "direct", 
               spline.types = "NCSS", df = NULL, lambdas = NULL, 
               smoothing.segments = list(c(11,20), c(21, 42)), 
               npspline.segments = NULL, 
               combinations = "allvalid")
args4smoothing(smoothing.methods = c("log","dir","log"), 
               spline.types = c("NCSS","NCSS","PS"), 
               df = c(4,5,NA), lambdas = c(NA,NA,0.36), 
              combinations = "parallel")
args4smoothing(smoothing.methods = "log", 
               spline.types = "PS", df = NULL, 
               lambdas = 0.36, combinations = "single")

Forms a smooths.frame from a data.frame, ensuring that the correct columns are present.

Description

Creates a smooths.frame from a data.frame by adding the class smooths.frame and a set of attributes to it.

Usage

as.smooths.frame(data, individuals = NULL, times = NULL)

Arguments

Value

A smooths.frame

Author(s)

Chris Brien

See Also

validSmoothsFrame, as.smooths.frame

Examples

dat <- read.table(header = TRUE, text = "
Type TunePar TuneVal Tuning Method       ID  DAP   PSA      sPSA
NCSS      df       4   df-4 direct 045451-C   28 57.446 51.18456
NCSS      df       4   df-4 direct 045451-C   30 89.306 87.67343
NCSS      df       7   df-7 direct 045451-C   28 57.446 57.01589
NCSS      df       7   df-7 direct 045451-C   30 89.306 87.01316
")
dat[1:7] <- lapply(dat[1:6], factor)
dat <- as.smooths.frame(dat, individuals = "ID", times = "DAP")
is.smooths.frame(dat)
validSmoothsFrame(dat)

Calculates the growth rates for a specified time interval for individuals in a data.frame in long format by taking weighted averages of growth rates for times within the interval.

Description

Using previously calculated growth rates over time, calculates the Absolute Growth Rates for a specified interval using the weighted averages of AGRs for each time point in the interval (AGR) and the Relative Growth Rates for a specified interval using the weighted geometric means of RGRs for each time point in the interval (RGR).

Usage

byIndv4Intvl_GRsAvg(data, responses, 
                    individuals = "Snapshot.ID.Tag", times = "DAP", 
                    which.rates = c("AGR","RGR"), 
                    suffices.rates=c("AGR","RGR"), sep.rates = ".",
                    start.time, end.time, 
                    suffix.interval, sep.suffix.interval = ".", 
                    sep.levels=".", na.rm=FALSE)

Arguments

Details

The AGR for an interval is calculated as the weighted mean of the AGRs for times within the interval. The RGR is calculated as the weighted geometric mean of the RGRs for times within the interval; in fact the exponential is taken of the weighted means of the logs of the RGRs. The weights are obtained from the times. They are taken as the sum of half the time subintervals before and after each time, except for the end points; the end points are taken to be the subintervals at the start and end of the interval.

Value

A data.frame with the growth rates. The name of each column is the concatenation of (i) one of responses, (ii) one of AGR, PGR or RGR, or the appropriate element of suffices.rates, and (iii) suffix.interval, the three components being separated by full stops.

Author(s)

Chris Brien

See Also

byIndv4Intvl_GRsDiff, byIndv4Intvl_WaterUse, getTimesSubset,
GrowthRates, byIndv4Times_SplinesGRs

Examples

data(exampleData)
longi.dat <- byIndv4Times_SplinesGRs(data = longi.dat, 
                                     response="PSA", response.smoothed = "sPSA", 
                                     individuals = "Snapshot.ID.Tag", 
                                     times = "DAP", 
                                     df = 4, 
                                     rates.method = "deriv", 
                                     which.rates = c("AGR", "RGR"), 
                                     suffices.rates = c("AGRdv", "RGRdv"))
sPSA.GR <- byIndv4Intvl_GRsAvg(data = longi.dat, 
                               response="sPSA", times = "DAP", 
                               which.rates = c("AGR","RGR"), 
                               suffices.rates = c("AGRdv","RGRdv"), 
                               start.time = 31, end.time = 35, 
                               suffix.interval = "31to35")

Calculates the growth rates for a specified time interval for individuals in a data.frame in long format by differencing the values for a response within the interval.

Description

Using the values of the responses, calculates the specified combination of the Absolute Growth Rates using differences (AGR), the Proportionate Growth Rates (PGR) and Relative Growth Rates using log differences (RGR) between two nominated time points.

Usage

byIndv4Intvl_GRsDiff(data, responses, 
                     individuals = "Snapshot.ID.Tag", times = "DAP", 
                     which.rates = c("AGR","PGR","RGR"), 
                     suffices.rates=NULL, sep.rates = ".", 
                     start.time, end.time, 
                     suffix.interval, sep.suffix.interval = ".")

Arguments

Details

The AGR is calculated as the difference between the values of response at the end.time and start.time divided by the difference between end.time and start.time. The PGR is calculated as the ratio of response at the end.time to that at start.time and the ratio raised to the power of the reciprocal of the difference between end.time and start.time. The RGR is calculated as the log of the PGR and so is equal to the difference between the logarithms of response at the end.time and start.time divided by the difference between end.time and start.time.

Value

A data.frame with the growth rates. The name of each column is the concatenation of (i) one of responses, (ii) one of AGR, PGR or RGR, or the appropriate element of suffices.rates, and (iii) suffix.interval, the three components being separated by full stops.

Author(s)

Chris Brien

See Also

byIndv4Intvl_GRsAvg, byIndv4Intvl_WaterUse, getTimesSubset, GrowthRates,
byIndv4Times_SplinesGRs

Examples

data(exampleData)
sPSA.GR <- byIndv4Intvl_GRsDiff(data = longi.dat, 
                                responses = "sPSA", times = "DAP", 
                                which.rates = c("AGR","RGR"), 
                                start.time = 31, end.time = 35,
                                suffix.interval = "31to35")

Calculates a single value that is a function of the values of an individual for a response in a data.frame in long format over a specified time interval.

Description

Splits the values of a response into subsets corresponding individuals and applies a function that calculates a single value from each individual's observations during a specified time interval. It includes the ability to calculate the observation number that is closest to the calculated value of the function and the assocated values of a factor or numeric.

Usage

byIndv4Intvl_ValueCalc(data, response, 
                       individuals = "Snapshot.ID.Tag", times = "DAP", 
                       FUN = "max", which.obs = FALSE, which.values = NULL, 
                       addFUN2name = TRUE, sep.FUNname = ".", 
                       start.time=NULL, end.time=NULL, 
                       suffix.interval=NULL, sep.suffix.interval = ".", 
                       sep.levels=".", weights=NULL, na.rm=TRUE, ...)

Arguments

Value

A data.frame, with the same number of rows as there are individuals, containing a column for the individuals and a column with the values of the function for the individuals. It is also possible to determine observaton numbers or the values of another column in data for the response values that are closest to the FUN results, using either or both of which.obs and which.values. If which.obs is TRUE, a column with observation numbers is included in the data.frame. If which.values is set to the name of a factor or a numeric, a column containing the levels of that factor or the values of that numeric is included in the data.frame.

The name of the column with the values of the function will be result of concatenating the response, FUN and, if it is not NULL, suffix.interval, each separated by a full stop. If which.obs is TRUE, the column name for the obervations numbers will have .obs added after FUN into the column name for the function values; if which.values is specified, the column name for these values will have a full stop followed by which.values added after FUN into the column name for the function values.

Author(s)

Chris Brien

See Also

byIndv4Intvl_GRsAvg, byIndv4Intvl_GRsDiff, byIndv4Intvl_WaterUse,
getTimesSubset

Examples

data(exampleData)
sPSA.max <- byIndv4Intvl_ValueCalc(data = longi.dat, 
                                   response = "sPSA", times = "DAP", 
                                   start.time = 31, end.time = 35, 
                                   suffix.interval = "31to35")
AGR.max.dat <- byIndv4Intvl_ValueCalc(data = longi.dat, 
                                      response = "sPSA", times = "DAP", 
                                      FUN="max", 
                                      start.time = 31, end.time = 35, 
                                      suffix.interval = "31to35",
                                      which.values = "DAP", 
                                      which.obs = TRUE)

Calculates water use traits (WU, WUR, WUI) over a specified time interval for each individual in a data.frame in long format.

Description

Calculates one or more of water use (WU), water use rate (WUR), and, for a set of responses, water use indices (WUI)s over a specified time interval for each individual in a data.frame in long format.

Usage

byIndv4Intvl_WaterUse(data, water.use = "Water.Use", responses = NULL, 
                      individuals = "Snapshot.ID.Tag", times = "DAP", 
                      trait.types = c("WU", "WUR", "WUI"), 
                      suffix.rate = "R", suffix.index = "I", 
                      sep.water.traits = "", sep.responses = ".", 
                      start.time, end.time, 
                      suffix.interval = NULL, sep.suffix.interval = ".", 
                      na.rm = FALSE)

Arguments

Details

WU

is the water use and is the sum of the water use after start.time until end.time. Thus, the water use up to start.time is not included. Further, if any water use value in the interval is NA, the interval value will be set to NA.

WUR

is the Water Use Rate and is WU divided by the difference between end.time and start.time.

WUI

is the Water Use Index and is calculated as a response difference between the start.time and the end.time, which is then divided by the WU.

Value

A data.frame containing the individuals column, WU and/or WUR and, if requested, a WUI for each element of responses. The names of WU and WUR will have suffix.interval appended, if it is not NULL, separated by a full stop (‘.’). The name of each WUI will be the concatenation of an element of responses with WUI and, if not NULL, suffix.interval, the three components being separated by a full stop (‘.’).

Author(s)

Chris Brien

See Also

byIndv4Intvl_GRsAvg, byIndv4Intvl_GRsDiff, getTimesSubset, GrowthRates

Examples

data(exampleData)
WU.WUI_31_35 <- byIndv4Intvl_WaterUse(data = longi.dat, 
                                      water.use = "WU", responses = "PSA", 
                                      times = "DAP", 
                                      trait.types = c("WUR","WUI"), 
                                      suffix.rate = ".Rate", 
                                      suffix.index = ".Index",
                                      start.time = 31, end.time = 35, 
                                      suffix.interval = "31to35")

Adds to a data.frame, the growth rates calculated for consecutive times for individuals in the data.frame in long format by differencing response values.

Description

Uses AGRdiff, PGR and RGRdiff to calculate growth rates continuously over time for the response by differencing pairs of pairs of response values and stores the results in data. The subsets are those values with the same levels combinations of the factors listed in individuals.

If avail.time.diffs is FALSE, the differences between consecutive time values are calculated. For this, it is assumed that the same first times value is present in data for all individuals.

Usage

byIndv4Times_GRsDiff(data, responses, 
                     individuals = "Snapshot.ID.Tag", times = "DAP", 
                     which.rates = c("AGR","PGR","RGR"), 
                     suffices.rates=NULL, sep.rates = ".", 
                     avail.times.diffs = FALSE, ntimes2span = 2)

Arguments

Value

A data.frame containing data to which has been added i) a column for the differences between the times, if it is not already in data, and (ii) columns with growth rates. The name of the column for times differences will be the value of times with ".diffs" appended. The name for each of the growth-rate columns will be either the value of response with one of ".AGR", ".PGR" or "RGR", or the corresponding value from suffices.rates appended. Each growth rate will be positioned at observation ceiling(ntimes2span + 1) / 2 relative to the two times from which the growth rate is calculated.

Author(s)

Chris Brien

See Also

smoothSpline, byIndv4Times_SplinesGRs

Examples

data(exampleData)
longi.dat <- byIndv4Times_GRsDiff(data = longi.dat, 
                                  response = "sPSA", 
                                  individuals = "Snapshot.ID.Tag", 
                                  times = "DAP", 
                                  which.rates=c("AGR", "RGR"), 
                                  avail.times.diffs = TRUE)

For a response in a data.frame in long format, computes, for a single set of smoothing parameters, smooths of the response, possibly along with growth rates calculated from the smooths.

Description

Uses smoothSpline to fit a spline to the values of response for each individual and stores the fitted values in data. The degree of smoothing is controlled by the tuning parameters df and lambda, related to the penalty, and by npspline.segments. The smoothing.method provides for direct and logarithmic smoothing.

The Absolute and Relative Growth Rates ( AGR and RGR) can be computed either using the first derivatives of the splines or by differencing the smooths. If using the first derivative to obtain growth rates, correctBoundaries must be FALSE. Derivatives other than the first derivative can also be produced. The function byIndv4Times_GRsDiff is used to obtain growth rates by differencing.

The handling of missing values in the observations is controlled via na.x.action and na.y.action. If there are not at least four distinct, nonmissing x-values, a warning is issued and all smoothed values and derivatives are set to NA.

The function probeSmooths can be used to investgate the effect the smoothing parameters
(smoothing.method, df or lambda) on the smooth that results.

Usage

byIndv4Times_SplinesGRs(data, response, response.smoothed = NULL, 
                        individuals = "Snapshot.ID.Tag", times, 
                        smoothing.method = "direct", smoothing.segments = NULL, 
                        spline.type = "NCSS", df=NULL, lambda = NULL, 
                        npspline.segments = NULL, 
                        correctBoundaries = FALSE, 
                        rates.method = "differences", 
                        which.rates = c("AGR","RGR"), 
                        suffices.rates = NULL, sep.rates = ".", 
                        avail.times.diffs = FALSE, ntimes2span = 2, 
                        extra.derivs = NULL, suffices.extra.derivs=NULL, 
                        sep.levels = ".", 
                        na.x.action="exclude", na.y.action = "trimx", ...)

Arguments

.

Value

A data.frame containing data to which has been added a column with the fitted smooth, the name of the column being the value of response.smoothed. If rates.method is not none, columns for the growth rates listed in which.rates will be added to data; the names each of these columns will be the value of response.smoothed with the elements of which.rates appended.

When rates.method is derivatives and smoothing.method is direct, the AGR is obtained from the first derivative of the spline for each value of times and the RGR is calculated as the AGR divided by the value of the response.smoothed for the corresponding time. When rates.method is derivatives and smoothing.method is logarithmic, the RGR is obtained from the first derivative of the spline and the AGR is calculated as the RGR multiplied by the corresponding value of the response.smoothed.

If extra.derivs is not NULL, the values for the nominated derivatives will also be added to data; the names each of these columns will be the value of response.smoothed with .dvf appended, where f is the order of the derivative, or the value of response.smoothed with the corresponding element of suffices.deriv appended.

Any pre-existing smoothed and growth rate columns in data will be replaced. The ordering of the data.frame for the times values will be preserved as far as is possible; the main difficulty is with the handling of missing values by the function merge. Thus, if missing values in times are retained, they will occur at the bottom of each subset of individuals and the order will be problematic when there are missing values in y and na.y.action is set to omit.

Author(s)

Chris Brien

References

Eilers, P.H.C and Marx, B.D. (2021) Practical smoothing: the joys of P-splines. Cambridge University Press, Cambridge.

Huang, C. (2001) Boundary corrected cubic smoothing splines. Journal of Statistical Computation and Simulation, 70, 107-121.

See Also

smoothSpline, probeSmooths, byIndv4Times_GRsDiff, smooth.spline, predict.smooth.spline,
split

Examples

data(exampleData)
#smoothing with growth rates calculated using derivates
longi.dat <- byIndv4Times_SplinesGRs(data = longi.dat, 
                                     response="PSA", response.smoothed = "sPSA", 
                                     times="DAP", 
                                     df = 4, rates.method = "deriv", 
                                     suffices.rates = c("AGRdv", "RGRdv"))
#Use P-splines
longi.dat <- byIndv4Times_SplinesGRs(data = longi.dat, 
                                     response="PSA", response.smoothed = "sPSA", 
                                     individuals = "Snapshot.ID.Tag", times="DAP", 
                                     spline.type = "PS", lambda = 0.1, 
                                     npspline.segments = 10, 
                                     rates.method = "deriv", 
                                     suffices.rates = c("AGRdv", "RGRdv"))
#with segmented smoothing and no growth rates
longi.dat <- byIndv4Times_SplinesGRs(data = longi.dat, 
                                     response="PSA", response.smoothed = "sPSA", 
                                     individuals = "Snapshot.ID.Tag", times="DAP", 
                                     smoothing.segments = list(c(28,34), c(35,42)), 
                                     df = 5, rates.method = "none")

Adds to a data.frame, water use traits calculated for consecutive times for individuals in the data.frame in long format.

Description

Calculates, for subsets of the data, those water traits out of Water Use (WU), Water Use Rate (WUR) and Water Use Index (WUI) that are specified in which.water.traits. The traits are calculated continuously over time for the water traits using the values of the variables for consecutive times stored in data. The weight.after column and whichever of weight.before and water.added is not NULL are used in the calculations. The subsets are those values with the same levels combinations of the factors listed in individuals.

If avail.time.diffs is FALSE, the differences between consecutive time values are calculated for each value of individuals.

Usage

byIndv4Times_WaterUse(data, weight.after = "Weight.After",
                    weight.before = NULL, water.added = NULL, 
                    responseAGR = NULL, 
                    individuals = "Snapshot.ID.Tag", times = "DAP", 
                    which.trait.types = c("WU","WUR","WUI"), 
                    water.trait.names = c("WU", "WUR", "WUI"),  
                    avail.times.diffs = FALSE, ntimes2span = 2)

Arguments

Details

If weight.before is not NULL, then the water use (WU) between consecutive times is calulated as value of weight.after for first time minus the weight.after for the second time. If water.added is not NULL, then the water use (WU) between consecutive times is calulated as value of the water.added minus the the difference between the weight.after values for the two consecutive times. The WUR is calculated as the WU divided by the times.diffs and the WUI is calculated as the ratio of the values on the column nomiated by responseAGR to the WUR values.

Value

A data.frame containing data to which has been added i) a column for the differences between the times, if it is not already in data, and (ii) columns with the water traits nominated in which.trait.types using the names specified in water.trait.names. The name of the column for times differences will be the value of times with ".diffs" appended. Each water trait will be positioned at observation ceiling(ntimes2span + 1) / 2 relative to the two times from which the growth rate is calculated.

Author(s)

Chris Brien

See Also

smoothSpline, byIndv4Times_GRsDiff

Examples

data(exampleData)
longi.dat <- byIndv4Times_WaterUse(data = longi.dat, 
                                   weight.before = "Weight.Before", 
                                   individuals = "Snapshot.ID.Tag", 
                                   times = "DAP", 
                                   which.trait.types = c("WU","WUR","WUI"), 
                                   water.trait.names = c("WU", "WUR","WUI"),
                                   responseAGR = "PSA.AGR")

Adds the necessary times and rates values to a data.frame to make a set of periodic, or equally spaced, rates.

Description

Adds the times and rates values to a data.frame to produce a set of periodic rates i.e. rates for equally spaced times, the difference between consecutive times being equal to reqd.times.diff. This assumes that the values for the consecutive differences between the values in the times column in data are an integer multiple of the value of reqd.times.diff and involves inserting times when the integer multiple of a times value is greater than one. It is also assumed that it is appropriate to use the rate value for the time immediately following inserted time(s) for the rate value(s) for the inserted times. Additional columns can be specified to have values duplicated for the inserted times.

If avail.time.diffs is FALSE, the differences between consecutive time values in data are calculated.

Usage

byIndv4Times_periodicRates(data, responses, 
                           individuals = "Snapshot.ID.Tag", 
                           times = "DAP", 
                           columns2duplicate = NULL, 
                           reqd.times.diff = 1, 
                           avail.times.diffs = FALSE)

Arguments

Value

A data.frame obtained by expanding data to include all combinations of the values of individuals and the set of equally spaced times between the minimum and maximun values of times, the spacing being the value of reqd.times.diff. For the rate(s) specified in the argument responses, values of the rate for the times immediately above an inserted times value are assigned to the values of the responses for the inserted times.

Author(s)

Chris Brien

See Also

byIndv4Times_WaterUse

Examples

data(exampleData)
longi.dat <- within(longi.dat, WUR <- WU/DAP.diffs)
longi.dat <- byIndv4Times_periodicRates(data = longi.dat, 
                                        responses = "WUR", 
                                        individuals = "Snapshot.ID.Tag", 
                                        times = "DAP", 
                                        avail.times.diffs = TRUE)

Calculates a single value that is a function of an individual's values for a response.

Description

Applies a function to calculate a single value from an individual's values for a response in a data.frame in long format. It includes the ability to calculate the observation number that is closest to the calculated value of the function and the assocated values of a factor or numeric.

Usage

byIndv_ValueCalc(data, response, individuals = "Snapshot.ID.Tag", 
                 FUN = "max", which.obs = FALSE, which.values = NULL, 
                 addFUN2name = TRUE, sep.FUNname = ".", 
                 weights = NULL, na.rm=TRUE, sep.levels = ".", ...)

Arguments

Value

A data.frame, with the same number of rows as there are individuals, containing a column for the individuals and a column with the values of the function for the individuals. It is also possible to determine observaton numbers or the values of another column in data for the response values that are closest to the FUN results, using either or both of which.obs and which.values. If which.obs is TRUE, a column with observation numbers is included in the data.frame. If which.values is set to the name of a factor or a numeric,a column containing the levels of that factor or the values of that numeric is included in the data.frame.

The name of the column with the values of the function will be formed by concatenating the response and FUN, separated by a full stop. If which.obs is TRUE, the column name for the obervations numbers will have .obs added after FUN into the column name for the function values; if which.values is specified, the column name for these values will have a full stop followed by which.values added after FUN into the column name for the function values.

Author(s)

Chris Brien

See Also

byIndv4Intvl_ValueCalc, byIndv4Times_GRsDiff, byIndv4Times_SplinesGRs

Examples

data(exampleData)
sPSA.max.dat <- byIndv_ValueCalc(data=longi.dat, response = "PSA")
AGR.max.dat <- byIndv_ValueCalc(data=longi.dat, response = "sPSA.AGR", 
                                FUN="max", 
                                which.values = "DAP", which.obs = TRUE)
sPSA.dec1.dat <- byIndv_ValueCalc(data=longi.dat, response = "sPSA", 
                                  FUN="quantile", 
                                  which.values = "DAP", probs = 0.1)

Replaces the values in a vector with the result of applying an operation to it and a lagged value

Description

Replaces the values in x with the result of applying an operation to it and the value that is lag positions either before it or after it in x, depending on whether lag is positive or negative. For positive lag the first lag values will be NA, while for negative lag the last lag values will be NA. When operation is NULL, the values are moved lag positions down the vector.

Usage

calcLagged(x, operation = NULL, lag = 1, ...)

Arguments

Value

A vector containing the result of applying operation to values in x. For positive lag the first lag values will be NA, while for negative lag the last lag values will be NA.

Author(s)

Chris Brien

See Also

Ops

Examples

data(exampleData)
longi.dat$DAP.diffs <- calcLagged(longi.dat$xDAP, operation ="-")

Calculates for a set of times, the time intervals after an origin time and the position of each within a time interval

Description

For the column specified in imageTimes, having converted it to POSIXct if not already converted, calculates for each value the number of intervalUnits between the time and the startTime. Then the number of timePositions within the intervals is calculated for the values in imageTimes. The function difftimes is used in doing the calculations, but the results are converted to numeric. For example intervals could correspond to the number of Days after Planting (DAP) and the timePositions to the hour within each day.

Usage

calcTimes(data, imageTimes = NULL, timeFormat = "%Y-%m-%d %H:%M",
          intervals = "Time.after.Planting..d.", startTime = NULL, 
          intervalUnit = "days", timePositions = NULL)

Arguments

Value

A data.frame, being the unchaged data data.frame when imageTimes is NULL or containing either intervals and/or timePositions depending on which is not NULL.

Author(s)

Chris Brien

See Also

as.POSIXct, imagetimesPlot.

Examples

  data(exampleData)
  raw.dat <- calcTimes(data = raw.dat, 
                       imageTimes = "Snapshot.Time.Stamp", timePositions = "Hour")

Calculates the cumulative sum, ignoring the first element if exclude.1st is TRUE

Description

Uses cumsum to calculate the cumulative sum, ignoring the first element if exclude.1st is TRUE.

Usage

cumulate(x, exclude.1st = FALSE, na.rm = FALSE, ...)

Arguments

Value

A vector containing the cumulative sum.

Author(s)

Chris Brien

See Also

cumsum

Examples

data(exampleData)
PSA.cum <- cumulate(longi.dat$PSA)

Adds the factors and covariates for a blocked, split-unit design

Description

Add the following factors and covariates to a date frame containing imaging data from the Plant Accelerator: Zone, xZone, SHZone, ZLane, ZMainunit, Subunit and xMainPosn. It checks that the numbers of levels of the factors are consistent with the observed numbers of individuals and measurements taken of them.

Usage

designFactors(data, insertName = NULL, designfactorMethod = "LanePosition", 
              nzones = 6, nlanesperzone = 4, 
              nmainunitsperlane = 11, nsubunitspermain = 2)

Arguments

Details

The factors Zone, ZLane, ZMainunit and Subunit are derived for each Smarthouse based on the values of nzones, nlanesperzone, nmainunitsperlane, nsubunitspermain, Zone being the blocks in the split-unit design. Thus, the number of individuals in each Smarthouse must be the product of these values and the number of observations must be the product of the numbers of smarthouse, indviduals and imagings for each individual. If this is not the case, it may be able to be achieved by including in data rows for extra observations that have values for the Snapshot.ID.Tag, Smarthouse, Lane, Position and Time.after.Planting..d. and the remaining columns for these rows have missing values (NA) Then SHZone is formed by combining Smarthouse and Zone and the covariates cZone, cMainPosn and cPosn calculated. The covariate cZone is calculated from Zone and cMainPosn is formed from the mean of cPosn for each main plot.

Value

A data.frame including the columns:

1. Smarthouse: factor with levels for the Smarthouse

2. Zone: factor dividing the Lanes into groups, usually of 4 lanes

3. cZone: numeric corresponding to Zone, centred by subtracting the mean of the unique positions

4. SHZone: factor for the combinations of Smarthouse and Zone

5. ZLane: factor for the lanes within a Zone

6. ZMainunit: factor for the main units within a Zone

7. Subunit: factor for the subunits

8. cMainPosn: numeric for the main-plot positions within a Lane, centred by subtracting the mean of the unique Positions

9. cPosn: numeric for the Positions within a Lane, centred by subtracting the mean of the unique Positions

Author(s)

Chris Brien

Examples

data(exampleData)
longi.dat <- prepImageData(data = raw.dat, smarthouse.lev = 1)
longi.dat <- designFactors(data = longi.dat, insertName = "Reps",
                           nzones = 1, nlanesperzone = 1, nmainunitsperlane = 10, 
                           designfactorMethod="StandardOrder")

A small data set to use in function examples

Description

Imaging data for 20 of the plants that were imaged over 14 days from an experiment in a Smarthouse in the Plant Accelerator. Producing these files is illustrated in the Rice vignette and the data is used as a small example in the growthPheno manual.

Usage

data(exampleData)

Format

Three data.frames:

1. raw.dat (280 rows by 34 columns) that contains the imaging data for 20 plants by 14 imaging days as produced by the image processing software;

2. longi.dat (280 rows by 37 columns) that contains a modified version of the imaging data for the 20 plants by 14 imaging days in raw.dat;

3. indv.dat (20 rows by 45 columns) that contains data summarizing the growth features of the 20 plants, based on the data in longi.dat.

Forms a subset of responses in data that contains their values for the nominated times

Description

Forms a subset of each of the responses in data that contains their values for the nominated times in a single column.

Usage

getTimesSubset(data, responses, 
               individuals = "Snapshot.ID.Tag", times = "DAP", 
               which.times, suffix = NULL, sep.suffix.times = ".", 
               include.times = FALSE, include.individuals = FALSE)

Arguments

Value

A data.frame containing the subset of responses ordered by as many of the initial columns of data as are required to uniquely identify each row (see order for more information). The names of the columns for each of the responses and for times in the subset are the concatenation of their names in data and suffix, separated by a full stop.

Author(s)

Chris Brien

Examples

data(exampleData)
sPSALast <- getTimesSubset("sPSA", data = longi.dat, times = "DAP", 
                           which.times = c(42), suffix = "last")

Deprecated Functions in the Package growthPheno

Description

These functions have been renamed and deprecated in growthPheno:

1. getDates -> getTimesSubset

2. anomPlot -> plotAnom

3. corrPlot -> plotCorrmatrix

4. fitspline -> smoothSpline

5. imagetimesPlot -> plotImagetimes

6. intervalGRaverage -> byIndv4Intvl_GRsAvg

7. intervalGRdiff -> byIndv4Intvl_GRsDiff

8. intervalValueCalculate -> byIndv4Intvl_ValueCalc

9. intervalWUI -> byIndv4Intvl_WaterUse

10. longiPlot -> plotProfiles

11. longitudinalPrime -> prepImageData

12. plotLongitudinal -> plotProfiles

13. plotMedianDeviations -> plotSmoothsMedianDevns

14. probeDF -> probeSmooths

15. probeSmoothing -> probeSmooths

16. splitContGRdiff -> byIndv4Times_GRsDiff

17. splitSplines -> byIndv4Times_SplinesGRs

18. splitValueCalculate -> byIndv4Intvl_ValueCalc

Usage

getDates(...)
anomPlot(...)
corrPlot(...)
fitSpline(...)
imagetimesPlot(...)
intervalGRdiff(...)
intervalGRaverage(...)
intervalValueCalculate(...)
intervalWUI(...)
longiPlot(...)
longitudinalPrime(...)
plotLongitudinal(...)
plotMedianDeviations(...)
probeDF(...)
probeSmoothing(...)
splitContGRdiff(...)
splitSplines(...)
splitValueCalculate(...)

Arguments

Author(s)

Chris Brien

Imports an Excel imaging file and allows some renaming of variables

Description

Uses readxl to import a sheet of imaging data produced by the Lemna Tec Scanalyzer. Basically, the data consists of imaging data obtained from a set of individuals (e.g. plant, pot, cart, plot or unit) over time. There should be a column, which by default is called Snapshot.ID.Tag, containing a unique identifier for each individual and a column, which by default is labelled Snapshot.Time.Stamp, containing the time of imaging for each observation in a row of the sheet. Also, if startTime is not NULL, calcTimes is called to calculate, or recalculate if already present, timeAfterStart from imageTimes by subtracting a supplied startTime.

Using cameraType, keepCameraType, labsCamerasViews and prefix2suffix, some flexibility is provided for renaming the columns with imaging data. For example, if the column names are prefixed with 'RGB_SV1', 'RGB_SV2' or 'RGB_TV', the 'RGB_' can be removed and the 'SV1', 'SV2' or 'TV' become suffices.

Usage

importExcel(file, sheet="raw data", sep = ",", 
            individualId = "Snapshot.ID.Tag", 
            imageTimes = "Snapshot.Time.Stamp", 
            timeAfterStart = "Time.after.Planting..d.", 
            cameraType = "RGB", keepCameraType = FALSE, 
            labsCamerasViews = NULL, prefix2suffix = TRUE, 
            startTime = NULL,
            timeFormat = "%Y-%m-%d %H:%M", 
            plotImagetimes = TRUE, ...)

Arguments

Value

A data.frame containing the data.

Author(s)

Chris Brien

See Also

as.POSIXct, calcTimes, plotImagetimes

Examples

filename <- system.file("extdata/rawdata.xlsx", package = "growthPheno", 
                        mustWork = TRUE)
raw.dat <- importExcel(file = filename, 
                       startTime  = "2015-02-11 0:00 AM")


camview.labels <- c("SF0", "SL0", "SU0", "TV0")
names(camview.labels) <- c("RGB_Side_Far_0", "RGB_Side_Lower_0", 
                           "RGB_Side_Upper_0", "RGB_TV_0")
filename <- system.file("extdata/raw19datarow.csv", package = "growthPheno", 
                        mustWork = TRUE)
raw.19.dat <- suppressWarnings(importExcel(file = filename,
                                           individualId = "Snapshot.ID.Tags",
                                           startTime = "06/10/2017 0:00 AM",
                                           timeFormat = "%d/%m/%Y %H:M", 
                                           labsCamerasViews = camview.labels, 
                                           plotImagetimes = FALSE))

Selects a subset of variables using Principal Variable Analysis (PVA), based on the observed values within a specified time interval

Description

Principal Variable Analysis (PVA) (Cumming and Wooff, 2007) selects a subset from a set of the variables such that the variables in the subset are as uncorrelated as possible, in an effort to ensure that all aspects of the variation in the data are covered. Here, all observations in a specified time interval are used for calculation the correlations on which the selection is based.

Usage

## S3 method for class 'data.frame'
intervalPVA(obj, responses, times = "Days", start.time, end.time, 
            nvarselect = NULL, p.variance = 1, include = NULL, 
            plot = TRUE, ...)

Arguments

Details

The variable that is most correlated with the other variables is selected first for inclusion. The partial correlation for each of the remaining variables, given the first selected variable, is calculated and the most correlated of these variables is selects for inclusion next. Then the partial correlations are adjust for the second included variables. This process is repeated until the specified criteria have been satisfied. The possibilities are to:

1. the default (nvarselect = NULL and p.variance = 1) select all variables in increasing order of amount of information they provide;

2. select exactly nvarselect variables;

3. select just enough variables, up to a maximum of nvarselect variables, to explain at least p.variance*100 per cent of the total variance.

Value

A data.frame giving the results of the variable selection. It will contain the columns Variable, Selected, h.partial, Added.Propn and Cumulative.Propn.

Author(s)

Chris Brien

References

Cumming, J. A. and D. A. Wooff (2007) Dimension reduction via principal variables. Computational Statistics and Data Analysis, 52, 550–565.

See Also

PVA, rcontrib

Examples

data(exampleData)
longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1)
longi.dat <- within(longi.dat, 
                    {
                      Max.Height <- pmax(Max.Dist.Above.Horizon.Line.SV1,  
                                         Max.Dist.Above.Horizon.Line.SV2)
                      Density <- PSA/Max.Height
                      PSA.SV = (PSA.SV1 + PSA.SV2) / 2
                      Image.Biomass = PSA.SV * (PSA.TV^0.5)
                      Centre.Mass <- (Center.Of.Mass.Y.SV1 + Center.Of.Mass.Y.SV2) / 2
                      Compactness.SV = (Compactness.SV1 + Compactness.SV2) / 2
                    })
responses <- c("PSA","PSA.SV","PSA.TV", "Image.Biomass", "Max.Height","Centre.Mass",
               "Density", "Compactness.TV", "Compactness.SV")
results <-  intervalPVA(longi.dat, responses, times = "DAP", 
                        start.time = "31", end.time = "31",
                        p.variance=0.9, plot = FALSE)

Tests whether an object is of class smooths.frame

Description

A single-line function that tests whether an object is of class smooths.frame.

Usage

is.smooths.frame(object)

Arguments

Value

A logical.

Author(s)

Chris Brien

See Also

validSmoothsFrame, as.smooths.frame

Examples

dat <- read.table(header = TRUE, text = "
Type TunePar TuneVal Tuning Method       ID  DAP   PSA      sPSA
NCSS      df       4   df-4 direct 045451-C   28 57.446 51.18456
NCSS      df       4   df-4 direct 045451-C   30 89.306 87.67343
NCSS      df       7   df-7 direct 045451-C   28 57.446 57.01589
NCSS      df       7   df-7 direct 045451-C   30 89.306 87.01316
")
dat[1:7] <- lapply(dat[1:7], factor)
dat <- as.smooths.frame(dat, individuals = "ID", times = "DAP")
is.smooths.frame(dat)
validSmoothsFrame(dat)

Identifies anomalous individuals and produces profile plots without them and with just them

Description

Uses byIndv4Intvl_ValueCalc and the function anom to identify anomalous individuals in longitudinal data. The user can elect to print the anomalous individuals, a profile plot without the anomalous individuals and/or a profile plot with only the anomalous individuals. The plots are produced using ggplot. The plot can be facettd so that a grid of plots is produced.

Usage

plotAnom(data, response="sPSA", 
         individuals="Snapshot.ID.Tag", 
         times = "DAP", x = NULL, 
         breaks.spacing.x = -2, angle.x = 0, 
         vertical.line=NULL, 
         groupsFactor=NULL, lower=NULL, upper=NULL, 
         start.time=NULL, end.time=NULL, 
         suffix.interval=NULL, 
         columns.retained=c("Snapshot.ID.Tag", "Smarthouse", "Lane", 
                            "Position", "Treatment.1", "Genotype.ID"),
         whichPrint=c("anomalous","innerPlot","outerPlot"), na.rm=TRUE, ...)

Arguments

Value

A list with three components:

1. data, a data frame resulting from the merge of data and the logical identifying whether or not an individual is anomalous;

2. innerPlot, an object of class ggplot storing the profile plot of the individuals that are not anomalous;

3. outerPlot, an object of class ggplot storing the profile plot of only the individuals that are anomalous.

The name of the column indicating anomalous individuals will be result of concatenating the response, anom and, if it is not NULL, suffix.interval, each separated by a full stop. The ggplot objects can be plotted using print and can be modified by adding ggplot functions before printing. If there are no observations to plot, NULL will be returned for the plot.

Author(s)

Chris Brien

See Also

anom, byIndv4Intvl_ValueCalc, ggplot2.

Examples

data(exampleData)
anomalous <- plotAnom(longi.dat, response="sPSA.AGR", 
                      times = "xDAP", 
                      lower=2.5, start.time=40, 
                      vertical.line=29, 
                      breaks.spacing.x = 2, 
                      whichPrint=c("innerPlot"), 
                      y.title="sPSA AGR")

Calculates and plots correlation matrices for a set of responses

Description

Having calculated the correlations a heat map indicating the magnitude of the correlations is produced using ggplot. In this heat map, the darker the red in a cell then the closer the correlation is to -1, while the deeper the blue in the cell, then the closer the correlation is to 1. Matrix plots of all pairwise combinations of the variables can be produced that includes the values of the the correlation coefficients. If pairs.sets is set, a matrix plot, along with the values of the correlation coefficients, is produced for each of the pair.sets. That is, the argument pairs.sets can be used to restrict the pairs in a matrix plot to those combinations within each set.

Usage

plotCorrmatrix(data, responses, which.plots = c("heatmap","matrixplots"), 
               title = NULL, labels = NULL, labelSize = 4, pairs.sets = NULL, 
               show.sig = TRUE, cell.text.size = 12, axis.text.size = 20, 
               ggplotFuncs = NULL, printPlot = TRUE, ...)

Arguments

Details

The correlations and their p-values are producced using rcorr from the Hmisc package. The heatmap is produced using ggplot and the matrixplots are produced using GGally.

Value

A list object that has components heatmap and matrixplots. The component heatmap will contain the heatmap plot, if produced, as an object of class "ggplot", which can be plotted using print; otherwise NULL is returned. Similarly, if not NULL, the component matrixplots will contain a list with one or more components, depending on the setting of pair.sets, each of which is a scatterplot matrix stored as an object of class "ggmatrix".

Author(s)

Chris Brien

See Also

rcorr, GGally, ggplot.

Examples

data(exampleData)
longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1)
longi.dat <- within(longi.dat, 
                    {
                      Max.Height <- pmax(Max.Dist.Above.Horizon.Line.SV1,  
                                         Max.Dist.Above.Horizon.Line.SV2)
                      Density <- PSA/Max.Height
                      PSA.SV = (PSA.SV1 + PSA.SV2) / 2
                      Image.Biomass = PSA.SV * (PSA.TV^0.5)
                      Centre.Mass <- (Center.Of.Mass.Y.SV1 + Center.Of.Mass.Y.SV2) / 2
                      Compactness.SV = (Compactness.SV1 + Compactness.SV2) / 2
                    })
responses <- c("PSA","PSA.SV","PSA.TV", "Image.Biomass", "Max.Height","Centre.Mass",
               "Density", "Compactness.TV", "Compactness.SV")
plotCorrmatrix(longi.dat, responses, pairs.sets=list(c(1:4),c(5:7)))

Produces boxplots of the deviations of the observed values from the smoothed values over values of x.

Description

Produces boxplots of the deviations of the observed values from the smoothed values over values of x.

Usage

plotDeviationsBoxes(data, observed, smoothed, x.factor, 
                    x.title = NULL, y.titles = NULL,
                    facet.x = ".", facet.y = ".", 
                    facet.labeller = NULL, 
                    facet.scales = "fixed", 
                    angle.x = 0, 
                    deviations.plots = "absolute", 
                    ggplotFuncs = NULL, printPlot = TRUE, ...)
 

Arguments

Value

A list whose components are named absolute and relative; a component will contain an object of class "ggplot" when the plot has been requested using the deviations.plots argument and a NULL otherwise. The objects can be plotted using print.

Author(s)

Chris Brien

See Also

plotSmoothsMedianDevns, probeSmooths, ggplot.

Examples

data(exampleData)

plotDeviationsBoxes(longi.dat, observed = "PSA", smoothed = "sPSA",
                    x.factor="DAP", facet.x.pf = ".", facet.y= ".", df =5)

Plots the position of a time within an interval against the interval for image

Description

Uses ggplot to produce a plot of the time position within an interval against the interval. For example, one might plot the hour of the day images are taken against the days after planting (or some other number of days after an event). A line is produced for each value of groupVariable and the colour is varied according to the value of the colourVariable. Each Smarthouse is plotted separately. It aids in checking whether delays occurred in imaging the plants.

Usage

plotImagetimes(data, intervals = "Time.after.Planting..d.", timePositions = "Hour", 
               groupVariable = "Snapshot.ID.Tag", colourVariable = "Lane", 
               ggplotFuncs = NULL, printPlot = TRUE)

Arguments

Value

An object of class "ggplot", which can be plotted using print.

Author(s)

Chris Brien

See Also

ggplot, calcTimes.

Examples

data(exampleData)
library(ggplot2)
longi.dat <- calcTimes(longi.dat, imageTimes = "Snapshot.Time.Stamp",
                       timePositions = "Hour")
plotImagetimes(data = longi.dat, intervals = "DAP", timePositions = "Hour",
               ggplotFuncs=list(scale_colour_gradient(low="grey20", high="black"), 
                                geom_line(aes(group=Snapshot.ID.Tag, colour=Lane))))

Produces profile plots of longitudinal data for a set of individuals

Description

Produce profile plots of longitudinal data for a response using ggplot. A line is drawn for the data for each individual and the plot can be faceted so that a grid of plots is produced. For each facet a line for the medians over time can be added, along with the vaue of the outer whiskers (median +/- 1.5 * IQR).

Usage

plotProfiles(data, response = "PSA", 
             individuals = "Snapshot.ID.Tag", times = "DAP", 
             x = NULL, title = NULL, 
             x.title = "DAP", y.title = "PSA (kpixels)", 
             facet.x = ".", facet.y = ".", 
             labeller = NULL, scales = "fixed", 
             breaks.spacing.x = -2, angle.x = 0, 
             colour = "black", 
             colour.column = NULL, colour.values = NULL, 
             alpha = 0.1, addMediansWhiskers = FALSE, 
             ggplotFuncs = NULL, 
             printPlot = TRUE)

Arguments

Value

An object of class "ggplot", which can be plotted using print.

Author(s)

Chris Brien

See Also

ggplot, labeller.

Examples

data(exampleData)
plotProfiles(data = longi.dat, response = "sPSA", times = "DAP")

plt <- plotProfiles(data = longi.dat, response = "sPSA", 
                    y.title = "sPSA (kpixels)", 
                    facet.x = "Treatment.1", facet.y = "Smarthouse", 
                    breaks.spacing.x = 2,
                    printPlot=FALSE)
plt <- plt + ggplot2::geom_vline(xintercept=29, linetype="longdash", linewidth=1) +
             ggplot2::scale_y_continuous(limits=c(0,750))
print(plt)

plotProfiles(data = longi.dat, response = "sPSA", times = "DAP", 
             x.title = "DAP", y.title = "sPSA (kpixels)", 
             facet.x = "Treatment.1", facet.y = "Smarthouse", 
                 ggplotFuncs = list(ggplot2::geom_vline(xintercept=29, 
                                                        linetype="longdash", 
                                                        size=1), 
                                    ggplot2::scale_x_continuous(breaks=seq(28, 42, 
                                                                           by=2)), 
                                    ggplot2::scale_y_continuous(limits=c(0,750))))

Plots several sets of smoothed values for a response, possibly along with growth rates and optionally including the unsmoothed values, as well as deviations boxplots.

Description

Plots the smoothed values for an observed response and, optionally, the unsmoothed observed response using plotProfiles. Depending on the setting of trait.types (response, AGR or RGR), the computed traits of the Absolute Growth Rates (AGR) and/or the Relative Growth Rates (RGR) are plotted. This function will also calculate and produce, using plotDeviationsBoxes, boxplots of the deviations of the supplied smoothed values from the observed response values for the traits and for combinations of the different smoothing parameters and for subsets of non-smoothing-factor combinations. The observed and smoothed values are supplied in long format i.e. with the values for each set of smoothing parameters stacked one under the other in the supplied smooths.frame. Such data can be generated using probeSmooths; to prevent probeSmooths producing the plots, which it is does using plotSmoothsComparison, plotDeviationsBoxes and plotSmoothsMedianDevns, set which.plots to none. The smoothing parameters include spline.types, df, lambdas and smoothing.methods (see probeSmooths).

Multiple plots, possibly each having multiple facets, are produced using ggplot2. The layout of these plots is controlled via the arguments plots.by, facet.x and facet.y. The basic principle is that the number of levels combinations of the smoothing-parameter factors Type, TunePar, TuneVal, Tuning (the combination of (TunePar and TuneVal), and Method that are included in plots.by, facet.x and facet.y must be the same as those covered by the combinations of the values supplied to spline.types, df, lambdas and Method and incorporated into the smooths.frame input to plotSmoothsComparison via the data argument. This ensures that smooths from different parameter sets are not pooled into the same plot. The factors other than the smoothing-parameter factors can be supplied to the plots.by and facet arguments.

The following profiles plots can be produced: (i) separate plots of the smoothed traits for each combination of the smoothing parameters (include Type, Tuning and Method in plots.by); (ii) as for (i), with the corresponding plot for the unsmoothed trait preceeding the plots for the smoothed trait (also set include.raw to alone); (iii) profiles plots that compare a smoothed trait for all combinations of the values of the smoothing parameters, arranging the plots side-by-side or one above the other (include Type, Tuning and Method in facet.x and/or facet.y - to include the unsmoothed trait set include.raw to one of facet.x or facet.y; (iv) as for (iii), except that separate plots are produced for each combination of the levels of the factors in plot.by and each plot compares the smoothed traits for the smoothing-parameter factors included in facet.x and/or facet.y (set both plots.by and one or more of facet.x and facet.y).

Usage

plotSmoothsComparison(data, response, response.smoothed = NULL, 
                      individuals = "Snapshot.ID.Tag", times = "DAP", 
                      trait.types = c("response", "AGR", "RGR"), 
                      x.title = NULL, y.titles = NULL, 
                      profile.plot.args = 
                        args4profile_plot(plots.by = NULL, 
                                          facet.x = ".", facet.y = ".", 
                                          include.raw = "no"),
                      printPlot = TRUE, ...)

Arguments

Value

A multilevel list that contains the ggplot objects for the plots produced. The first-level list has a component for each trait.types and each of these is a second-level list that contains the trait profile plots and for a trait. It may contain components labelled Unsmoothed, all or for one of the levels of the factors in plots.by; each of these third-level ists contains a ggplot object that can be plotted using print.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, args4profile_plot, plotDeviationsBoxes, plotSmoothsMedianDevns, ggplot2.

Examples

data(exampleData)
vline <- list(ggplot2::geom_vline(xintercept=29, linetype="longdash", size=1))
traits <- probeSmooths(data = longi.dat, 
                       response = "PSA", response.smoothed = "sPSA", 
                       times = "DAP", 
                       #only df is changed from the probeSmooth default
                       smoothing.args = 
                         args4smoothing(smoothing.methods = "direct", 
                                        spline.types = "NCSS", 
                                        df = c(4,7), lambdas = NULL), 
                       which.plots = "none")
plotSmoothsComparison(data = traits, 
                      response = "PSA", response.smoothed = "sPSA", 
                      times = "DAP", x.title = "DAP",
                      #only facet.x is changed from the probeSmooth default
                      profile.plot.args = 
                        args4profile_plot(plots.by = NULL, 
                                          facet.x = "Tuning", facet.y = ".", 
                                          include.raw = "no", 
                                          ggplotFuncs = vline))

Produces boxplots for several sets of deviations of the smoothed values from a response, possibly along with growth rates.

Description

Calculates and produces, using plotDeviationsBoxes, boxplots of the deviations of the supplied smoothed values from the observed response values for the traits and for combinations of the different smoothing parameters and for subsets of non-smoothing-factor combinations. Which traits are plotted is controlled by trait.types and may include the (responseand the computed traits of the Absolute Growth Rates (AGR) and/or the Relative Growth Rates (RGR). The observed and smoothed values are supplied in long format i.e. with the values for each set of smoothing parameters stacked one under the other in the supplied smooths.frame. Such data can be generated using probeSmooths.

Multiple plots, possibly each having multiple facets, are produced using ggplot2. The layout of these plots is controlled via the arguments plots.by, facet.x and facet.y. The basic principle is that the number of levels combinations of the smoothing-parameter factors Type, TunePar, TuneVal, Tuning (the combination of (TunePar and TuneVal), and Method that are included in plots.by, facet.x and facet.y must be the same as those covered by the combinations of the values incorporated into the smooths.frame input to plotSmoothsDevnBoxplots via the data argument. This ensures that smooths from different parameter sets are not pooled into the same plot. The factors other than the smoothing-parameter factors can be supplied to the plots.by and facet arguments.

Usage

plotSmoothsDevnBoxplots(data, response, response.smoothed = NULL, 
                        individuals = "Snapshot.ID.Tag", times = "DAP", 
                        trait.types = c("response", "AGR", "RGR"), 
                        which.plots =  "absolute.boxplots", 
                        x.title = NULL, y.titles = NULL, 
                        devnboxes.plot.args = 
                          args4devnboxes_plot(plots.by = NULL, 
                                              facet.x = ".", facet.y = "."),
                        printPlot = TRUE, ...)

Arguments

Value

A multilevel list that contains the ggplot objects for the plots produced. The first-level list has a component for each trait.types and each of these is a second-level list with contains the deviations boxplots for a response. Each plot is in an object of class ggplot, which can be plotted using print.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, args4profile_plot, plotDeviationsBoxes, plotSmoothsMedianDevns, ggplot.

Examples

data(exampleData)
traits <- probeSmooths(data = longi.dat, 
                       response = "PSA", response.smoothed = "sPSA", 
                       times = "DAP", 
                       #only df is changed from the probeSmooth default
                       smoothing.args = 
                         args4smoothing(smoothing.methods = "direct", 
                                        spline.types = "NCSS", 
                                        df = c(4,7), lambdas = NULL), 
                       which.plots = "none")
plotSmoothsDevnBoxplots(data = traits, 
                        response = "PSA", response.smoothed = "sPSA", 
                        times = "DAP", x.title = "DAP",
                        #only facet.x is changed from the probeSmooth default
                        devnboxes.plot.args = 
                          args4devnboxes_plot(plots.by = NULL, 
                                              facet.x = "Tuning", facet.y = "."))

Calculates and plots the medians of the deviations from the observed values for several sets of smoothed values stored in a data.frame in long format.

Description

Calculates and plots the medians of the deviations of the supplied smoothed values from the supplied observed values for traits and combinations of different smoothing parameters, possibly for subsets of non-smoothing-factor combinations. The observed and smoothed values are supplied in long format i.e. with the values for each set of smoothing parameters stacked one under the other in the supplied data.frame. Such data can be generated using probeSmooths; to prevent probeSmooths producing the plots, which it is does using plotSmoothsComparison, plotDeviationsBoxes and plotSmoothsMedianDevns, set which.plots to none. The smoothing parameters include spline.types, df, lambdas and smoothing.methods (see probeSmooths).

Multiple plots, possibly each having multiple facets, are produced using ggplot2. The layout of these plots is controlled via the smoothing-parameter factors Type, Tuning (the combination of TunePar and TuneVal) and Method that can be supplied to the arguments plots.by, plots.group, facet.x and facet.y. These plots and facet arguments can also include factors other than the smoothing-parameter factors, that are also associated with the data. The basic principle is that the number of levels combinations of the smoothing-parameter factors included in the plots and facet arguments must be the same as those covered by the combinations of the values supplied to spline.types, df, lambdas and Method and incorporated into the smooths.frame input to plotSmoothsMedianDevns via the data argument. This ensures that smooths from different parameter sets are not pooled in a single plot. Envelopes of the median value of a trait for each factor combination can be added.

Usage

plotSmoothsMedianDevns(data, response, response.smoothed = NULL, 
                       individuals = "Snapshot.ID.Tag",  times = "DAP", 
                       trait.types = c("response", "AGR", "RGR"), 
                       x.title = NULL, y.titles = NULL, 
                       meddevn.plot.args = 
                         args4meddevn_plot(plots.by = NULL, plots.group = NULL,
                                           facet.x = ".", facet.y = ".", 
                                           propn.note = TRUE, 
                                           propn.types = c(0.1, 0.5, 0.75)), 
                       printPlot = TRUE, ...)

Arguments

Value

A list that consists of two components: (i) a componenent named plots that stores a two-level list of the median deviations plots; the first-level list has a component for each trait.types and each of these list(s) is a second-level list that contains the set of plots specified by plots.by (if plots.by is NULL, a single plot is stored); (ii) a component named med.dev.dat that stores the data.frame containing the median deviations that have been plotted. Each plot in the plots list is in an object of class ggplot, which can be plotted using print.

Author(s)

Chris Brien

See Also

traitSmooth, probeSmooths, args4meddevn_plot, plotSmoothsComparison, plotDeviationsBoxes, ggplot.

Examples

data(exampleData)
vline <- list(ggplot2::geom_vline(xintercept=29, linetype="longdash", size=1))
traits <- probeSmooths(data = longi.dat, 
                       response = "PSA", response.smoothed = "sPSA", 
                       times = "DAP", 
                       get.rates = FALSE, trait.types = "response", 
                       smoothing.args = 
                         args4smoothing(smoothing.methods = "direct", 
                                        spline.types = "NCSS", 
                                        df = c(4,7), lambdas = NULL), 
                       which.plots = "none")
med <- plotSmoothsMedianDevns(data = traits, 
                              response = "PSA", response.smoothed = "sPSA", 
                              times = "DAP", trait.types = "response",  
                              meddevn.plot.args = 
                                args4meddevn_plot(plots.by = NULL, 
                                                  plots.group = "Tuning", 
                                                  facet.x = ".", facet.y = ".", 
                                                  propn.types = 0.02,
                                                  ggplotFuncs = vline))

Prepares raw imaging data for further processing

Description

Forms the prime traits by selecting a subset of the traits in a data.frame of imaging data produced by the Lemna Tec Scanalyzer. The imaging traits to be retained are specified using the traits and labsCamerasViews arguments. Some imaging traits are divided by 1000 to convert them from pixels to kilopixels. Also added are factors and explanatory variates that might be of use in an analysis of the data.

Usage

prepImageData(data, individualId = "Snapshot.ID.Tag", 
              imageTimes = "Snapshot.Time.Stamp", 
              timeAfterStart = "Time.after.Planting..d.",
              PSAcolumn = "Projected.Shoot.Area..pixels.", 
              potIDcolumns = NULL, 
              idcolumns = c("Genotype.ID","Treatment.1"), 
              traits = list(all = c("Area", 
                                    "Boundary.Points.To.Area.Ratio", 
                                    "Caliper.Length", "Compactness", 
                                    "Convex.Hull.Area"), 
                            side = c("Center.Of.Mass.Y", 
                                     "Max.Dist.Above.Horizon.Line")),
              labsCamerasViews = list(all = c("SV1", "SV2", "TV"),
                                      side = c("SV1", "SV2")), 
              smarthouse.lev = NULL, 
              calcWaterUse = TRUE, ...)

Arguments

Details

The columns are copied from data, except for those columns that are calculated from the columns in data; those columns that are calculated have ‘(calculated)’ appended in the list under Value.

Value

A data.frame containing the columns specified by individualId, imageTimes, timeAfterStart, potIDcolumns or, if potIDcolumns is NULL, idcolumns, traits and cameras. The defaults will result in the following columns:

1. Smarthouse: factor with levels for the Smarthouse

2. Lane: factor for lane number in a smarthouse

3. Position: factor for east/west position in a lane

4. DAP: factor for the number of Days After Planting

5. xDAP: numeric for the DAP (calculated)

6. individualId: unique code for each individual

7. imageTimes: time at which an image was taken in POSIXct format

8. Hour: hour of the day, to 2 decimal places, at which the image was taken (calculated)

9. potIDcolumns: the columns listed in potIDcolumns, after being converted to factors

10. Reps: factor indexing the replicates for each combination of the factors in idcolumns (calculated only if potIDcolumns is NULL)

11. idcolumns: only if potIDcolumns is NULL, the columns listed in idcolumns, after being converted to factors

12. Weight.Before: weight of the pot before watering (only if calcWaterUse is TRUE)

13. Weight.After: weight of the pot after watering (only if calcWaterUse is TRUE)

14. Water.Amount: the weight of the water added (= Water.After - Water.Before) (calculated)

15. WU: the water use calculated as the difference between Weight.Before for the current imaging and the Weight.After for the previous imaging (calculated unless calcWaterUse is FALSE)

16. PSA: the Projected.Shoot.Area..pixels. divided by 1000 (calculated)

17. PSA.SV1: the Projected.Shoot.Area from Side View 1 divided by 1000 (calculated)

18. PSA.SV2: the Projected.Shoot.Area from Side View 2 divided by 1000 (calculated)

19. PSA.TV: the Projected.Shoot.Area from Top View divided by 1000 (calculated)

20. Boundary.To.PSA.Ratio.SV1

21. Boundary.To.PSA.Ratio.SV2

22. Boundary.To.PSA.Ratio.TV

23. Caliper.Length.SV1

24. Caliper.Length.SV2

25. Caliper.Length.TV

26. Compactness.SV1 from Side View 1

27. Compactness.SV2 from Side View 2

28. Compactness.TV: from Top View

29. Convex.Hull.PSA.SV1: area of Side View 1 Convex Hull divided by 1000 (calculated)

30. Convex.Hull.PSA.SV2: area of Side View 2 Convex Hull divided by 1000 (calculated)

31. Convex.Hull.PSA.TV: Convex.Hull.Area.TV divided by 1000 (calculated)

32. Center.Of.Mass.Y.SV1: Centre of Mass from Side View 1

33. Center.Of.Mass.Y.SV2: Centre of Mass from Side View 2

34. Max.Dist.Above.Horizon.Line.SV1: the Max.Dist.Above.Horizon.Line.SV1 divided by 1000 (calculated)

35. Max.Dist.Above.Horizon.Line.SV2: the Max.Dist.Above.Horizon.Line.SV2 divided by 1000 (calculated)

Author(s)

Chris Brien

Examples

data(exampleData)
longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1)

longi.dat <- prepImageData(data=raw.dat, 
                          traits= list(all = "Area"), 
                          labsCamerasViews = list(all = NULL), 
                          potIDcolumns = c("Genotype.ID", "Treatment.1", "Replicate" ), 
                          smarthouse.lev=1)

longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1, 
                           traits = list(a = "Area", c = "Compactness"),
                           labsCamerasViews = list(all = c("SV1", "SV2", "TV"), 
                                                   t = "TV"))

longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1, 
                           traits = c("Area.SV1", "Area.SV2", "Area.TV", 
                                      "Compactness.TV"),
                           labsCamerasViews = NULL)

longi.dat <- prepImageData(data=raw.dat, smarthouse.lev=1, 
                           calcWaterUse = FALSE, 
                           traits = list(img = c("Area", "Compactness"), 
                                         H20 = c("Weight.Before","Weight.After",
                                                 "Water.Amount")),
                          labsCamerasViews = list(all = c("SV1", "SV2", "TV"), 
                                                       H2O = NULL))

For a response in a data.frame in long format, computes and compares, for sets of smoothing parameters, smooths of the response, possibly along with growth rates calculated from the smooths.

Description

Takes an observed response and, for each individual, uses byIndv4Times_SplinesGRs to smooth its values employing the smoothing parameters specified by (i) spline.types, (ii) the tuning parameters, being the degrees of freedom values in df or the smoothing penalties in lambdas, and (iii) the smoothing.methods. The values of these, and other, smoothing arguments are set using the helper function args4smoothing.

Provided get.rates is TRUE or includes raw and/or smoothed and depending on the setting of trait.types, the Absolute Growth Rates (AGR) and/or the Relative Growth Rates (RGR) are calculated for each individual from the unsmoothed, observed response and from the smooths of the response, using either differences or first derivatives, as specified by rates.method.

Generally, profile plots for the traits (a response, an AGR or an RGR) specified in traits.types are produced if which.plots is profiles; if which.plots specifies one or more deviations plots, then those deviations plots will also be produced, these being based on the unsmoothed data from which the smoothed data has been subtracted. The layout of the plots is controlled via combinations of one or more of the smoothing-parameter factors Type, TunePar, TuneVal, Tuning (the combination of TunePar and TuneVal) and Method, as well as other factors associated with the data. The factors that are to be used for the profile plots are supplied via the argument profile.plot.args using the helper function args4profile_plot and for the and deviations boxplots using the helper function args4devnboxes_plot. These helper functions set plots.by, facet.x, and facet.y. For the plots of the medians of the deviations, the factors are supplied via the argument meddevn.plot.args using the helper function args4meddevn_plot to set plots.by, facet.x, facet.y and plots.group. Here, the basic principle is that the number of levels combinations of the smoothing-parameter factors included in the set of plots and facets arguments to one of these helper functions must be the same as those covered by the combinations of the values supplied to spline.types, df, lambdas and smoothing.methods and incorporated into the smooths.frame, such as is returned by probeSmooths. This ensures that smooths from different parameter sets are not pooled together in a single plot. It is also possible to include factors that are not smoothing-parameter factors in the plots amd facets arguments.

The following profiles plots can be produced using args4profile_plot: (i) separate plots of the smoothed traits for each combination of the smoothing parameters (include Type, Tuning and Method in plots.by); (ii) as for (i), with the corresponding plot for the unsmoothed trait preceeding the plots for the smoothed trait (also set include.raw to alone); (iii) profiles plots that compare a smoothed trait for all combinations of the values of the smoothing parameters, arranging the plots side-by-side or one above the other (include Type, Tuning and Method in facet.x and/or facet.y - to include the unsmoothed trait set include.raw to one of facet.x or facet.y; (iv) as for (iii), except that separate plots are produced for each combination of the levels of the factors in plot.by and each plot compares the smoothed traits for the smoothing-parameter factors included in facet.x and/or facet.y (set both plots.by and one or more of facet.x and facet.y).

Deviation plots that can be produced are the absolute and relative deviations boxplots and plots of medians deviations (see which.plots).

The handling of missing values is controlled via na.x.action and na.y.action supplied to the helper function args4smoothing.

The probeSmooths arguments are grouped according to function in the following order:

1. Data description arguments: data, response, response.smoothed, individuals, times, keep.columns, trait.types, get.rates, rates.method, ntimes2span.

2. Smoothing arguments: smoothing.args (see args4smoothing).

3. General plot control: x.title, y.titles, facet.labeller, which.plots.

4. Profile plots (pf) features: profile.plot.args (see args4profile_plot)

5. Median-deviations (med) plots features: meddevn.plot.args (see args4meddevn_plot)

6. Deviations boxplots (box) features: devnboxes.plot.args (see args4devnboxes_plot)

Usage

probeSmooths(data, response = "PSA", response.smoothed = NULL, 
             individuals="Snapshot.ID.Tag", times = "DAP", 
             keep.columns = NULL, 
             get.rates = TRUE, 
             rates.method="differences", ntimes2span = NULL, 
             trait.types = c("response", "AGR", "RGR"), 
             smoothing.args = 
               args4smoothing(smoothing.methods = "direct", 
                              spline.types = "NCSS", 
                              df = NULL, lambdas = NULL), 
             x.title = NULL, y.titles = NULL, 
             which.plots = "profiles", 
             profile.plot.args = 
               args4profile_plot(plots.by = NULL, 
                                 facet.x = ".", facet.y = ".", 
                                 include.raw = "no"), 
             meddevn.plot.args = 
               args4meddevn_plot(plots.by = NULL, plots.group = NULL, 
                                 facet.x = ".", facet.y = ".",
                                 propn.note = TRUE, 
                                 propn.types = c(0.1, 0.5, 0.75)), 
             devnboxes.plot.args = 
               args4devnboxes_plot(plots.by = NULL, 
                                 facet.x = ".", facet.y = ".", 
                                 which.plots = "none"), 
             ...)

Arguments

Value

A smooths.frame that contains the unsmoothed and smoothed data in long format. That is, all the values for either an unsmoothed or a smoothed trait are in a single column. The smooths for a trait for the different combinatons of the smoothing parameters are placed in rows one below the other. The columns that are included in the smooths.frame are Type, TunePar, TuneVal, Tuning and Method, as well as those specified by individuals, times, response, and response.smoothed. and any included in the keep.columns, plots and facet arguments. If trait.types includes AGR or RGR, then the included growth rate(s) of the response and response.smoothed must be present, unless get.rates is TRUE or includes raw and/or smoothed. In this case, the growth rates specified by trait.types will be calculated for the responses nominated by get.rates and the differences between the times used in calculating the rates will be computed and added. Then, the names of the growth rates are formed from response and response.smoothed by appending .AGR and .RGR as appropriate; the name of the column with the times differences will be formed by appending .diffs to the value of times. The external.smooths will also be included. A smooths.frame has the attributes described in smooths.frame.

Columns in the supplied data.frame that have not been used in probeSmooths will not be included in the returned smooths.frame. If they might be needed subsequently, such as when extra plots are produced, they can be included in the smooths.frame by listing them in a character vector for the keep.columns argument.

The smooths.frame is returned invisibly.

Author(s)

Chris Brien

See Also

args4smoothing, , args4meddevn_plot, args4profile_plot, traitSmooth,
smoothSpline, byIndv4Times_SplinesGRs, byIndv4Times_GRsDiff, smooth.spline, psNormal, plotSmoothsComparison,
plotSmoothsMedianDevns, ggplot.

Examples

data(exampleData)
longi.dat <- longi.dat[1:140,] #reduce to a smaller data set
vline <- list(ggplot2::geom_vline(xintercept=29, linetype="longdash", linewidth=1))
yfacets <- c("Smarthouse", "Treatment.1")
probeSmooths(data = longi.dat, 
             response = "PSA", response.smoothed = "sPSA",
             individuals = "Snapshot.ID.Tag",times = "DAP", 
             smoothing.args = 
               args4smoothing(df = c(4,7), 
                              lambda = list(PS = c(0.316,10))),
             profile.plot.args = 
               args4profile_plot(plots.by = NULL, 
                                 facet.x = "Tuning", 
                                 facet.y = c("Smarthouse", "Treatment.1"), 
                                 include.raw = "no", 
                                 alpha  = 0.4, 
                                 colour.column = "Method", 
                                 colour.values = c("orange", "olivedrab"), 
                                 ggplotFuncs = vline))

#An example that supplies three smoothing schemes to be compared
data(tomato.dat)
probeSmooths(data = tomato.dat, 
             response = "PSA", response.smoothed = "sPSA", 
             times = "DAP", 
             smoothing.args = 
               args4smoothing(spline.types      = c(  "N", "NCS",   "P"), 
                              df                = c(   4,     6,    NA), 
                              lambdas           = c(  NA,    NA,     1), 
                              smoothing.methods = c("dir", "log", "log"),
                              combinations      = "parallel"),
             which.plots = "medians.deviations",
             meddevn.plot.args = 
               args4meddevn_plot(plots.by = NULL, 
                                 plots.group = c("Type", "Tuning", "Method"),              
                                 facet.x = ".", facet.y = ".", 
                                 propn.note = FALSE, propn.types = NULL))

Computes a measure of how correlated each variable in a set is with the other variable, conditional on a nominated subset of them

Description

A measure of how correlated a variable is with those in a set is given by the square root of the sum of squares of the correlation coefficients between the variables and the other variables in the set (Cumming and Wooff, 2007). Here, the partial correlation between the subset of the variables listed in response that are not listed in include is calculated from the partial correlation matrix for the subset, adjusting for those variables in include. This is useful for manually deciding which of the variables not in include should next be added to it.

Usage

rcontrib(obj, ...)

Arguments

Details

rcontrib is the generic function for the rcontrib method. Use methods("rcontrib") to get all the methods for the rcontrib generic.

rcontrib.data.frame is a method for a data.frame.