Package 'leafareaR'

Title: Leaf Area Modeling, Evaluation, and Prediction
Description: Tools for leaf area estimation based on leaf length, leaf width, and observed leaf area. The package supports data validation, predictor generation, descriptive statistics, exploratory graphics, scatterplot matrices, linear models, nonlinear models, mixed models, model evaluation, ranking, equation generation, prediction, export of results and plots, and an interactive 'shiny' application. Methods implemented in the package are aligned with non-destructive allometric workflows described by Ribeiro et al. (2024) <doi:10.1016/j.sajb.2024.07.006>, Ribeiro et al. (2023) <doi:10.1590/1807-1929/agriambi.v27n3p209-215>, and Ribeiro et al. (2025) <doi:10.1590/0103-8478cr20230550>.
Authors: Joao Everthon da Silva Ribeiro [aut, cre], Ester dos Santos Coelho [aut], Toshik Iarley da Silva [aut]
Maintainer: Joao Everthon da Silva Ribeiro <[email protected]>
License: GPL (>= 3)
Version: 0.0.1
Built: 2026-07-07 08:17:34 UTC
Source: https://github.com/agrobioestat/leafarear

Help Index

Calculate absolute bias

Description

Calculate absolute bias

Usage

la_abs_bias_metric(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_abs_bias_metric(obs, pred)

Add equation information to a results table

Description

Adds equation, coefficients_text, and random_effect columns to the results or summary table returned by a model-fitting function.

Usage

la_add_equation_to_results(fit_object, digits = 4)

Arguments

fit_object

A fit object containing models and either results or summary.
digits

Number of decimal places used in displayed equations.

Value

A data.frame.

Calculate prediction bias

Description

Calculate prediction bias

Usage

la_bias(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_bias(obs, pred)

Build a readable equation from a fitted model

Description

Returns a readable equation starting with ⁠LA =⁠.

Usage

la_build_equation(model, digits = 4)

Arguments

model

A fitted model object.
digits

Number of decimal places used in the displayed equation.

Value

A character string.

Calculate Lin's concordance correlation coefficient

Description

Calculate Lin's concordance correlation coefficient

Usage

la_ccc(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_ccc(obs, pred)

Create derived leaf parameters

Description

Generates derived variables from leaf length (L) and leaf width (W) to support leaf area (LA) modeling.

Usage

la_create_derived(data, variables = "all", keep_original = TRUE)

Arguments

data

A data.frame containing at least L and W.
variables

Character vector with the derived variables to create. Use "all" to create all available derived variables.
keep_original

Logical. If TRUE, keeps original columns in the output.

Value

A data.frame with derived variables added.

Examples

data(leafarea_sample)
head(la_create_derived(leafarea_sample, variables = c("LW", "L2", "W2")))

Calculate Willmott's index of agreement

Description

Calculate Willmott's index of agreement

Usage

la_d(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_d(obs, pred)

Summarize the default leaf area variables

Description

Returns descriptive statistics for the default variables available in a leafareaR dataset, using any combination of L, W, and LA found in the supplied data.frame.

Usage

la_descriptive_default(data, na.rm = TRUE, digits = 4)

Arguments

data

A data.frame containing at least one of L, W, or LA.
na.rm

Logical. If TRUE, missing values are removed.
digits

Number of decimal places used to round the output.

Value

A data.frame with descriptive statistics for the available default variables.

Examples

data(leafarea_sample)
la_descriptive_default(leafarea_sample)

Calculate descriptive statistics for selected variables

Description

Computes a descriptive summary for selected numeric variables in a leaf area dataset. The output is intended to support data exploration before model fitting and can be applied to the original measurements as well as derived predictors.

Usage

la_descriptive_stats(data, variables = NULL, na.rm = TRUE, digits = 4)

Arguments

data

A data.frame containing numeric variables.
variables

Character vector with variable names to summarize. If NULL, all numeric variables in data are used.
na.rm

Logical. If TRUE, missing values are removed before computing statistics.
digits

Number of decimal places used to round the output.

Value

A data.frame with descriptive statistics for each selected variable.

Examples

data(leafarea_sample)
la_descriptive_stats(leafarea_sample, variables = c("L", "W", "LA"))

Evaluate all linear models from a la_fit_linear_models() object

Description

Evaluate all linear models from a la_fit_linear_models() object

Usage

la_evaluate_linear_models(fit_object, digits = 4)

Arguments

fit_object

Object returned by la_fit_linear_models().
digits

Number of decimal places for rounding.

Value

A data.frame with metrics for all candidate linear models.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
met[, c("model_id", "RMSE", "R2")]

Evaluate all mixed models from a la_fit_mixed_models() object

Description

Evaluate all mixed models from a la_fit_mixed_models() object

Usage

la_evaluate_mixed_models(fit_object, digits = 4)

Arguments

fit_object

Object returned by la_fit_mixed_models().
digits

Number of decimal places for rounding.

Value

A data.frame with metrics for all candidate mixed models.

Examples

data(leafarea_sample)
fit <- la_fit_mixed_models(leafarea_sample, group_var = "species")
la_evaluate_mixed_models(fit)[, c("model_id", "RMSE", "R2")]

Evaluate a single fitted model

Description

Evaluate a single fitted model

Usage

la_evaluate_model(
  model,
  data = NULL,
  response = "LA",
  model_id = NA_character_,
  model_type = NA_character_,
  digits = 4
)

Arguments

model

A fitted model object (lm, nls, or merMod).
data

Optional data.frame used to fit the model. Required for nls and mixed models.
response

Character string with the response variable name.
model_id

Optional model identifier.
model_type

Optional model type label.
digits

Number of decimal places for rounding.

Value

A one-row data.frame.

Examples

data(leafarea_sample)
dat <- la_create_derived(leafarea_sample, variables = "LW")
m <- lm(LA ~ LW, data = dat)
la_evaluate_model(m, model_id = "lm_LW", model_type = "linear")

Evaluate all nonlinear models from a la_fit_nonlinear_models() object

Description

Evaluate all nonlinear models from a la_fit_nonlinear_models() object

Usage

la_evaluate_nonlinear_models(fit_object, digits = 4)

Arguments

fit_object

Object returned by la_fit_nonlinear_models().
digits

Number of decimal places for rounding.

Value

A data.frame with metrics for all candidate nonlinear models.

Examples

data(leafarea_sample)
fit <- la_fit_nonlinear_models(leafarea_sample, 
                               models = c("power_LW", "exponential_LW"))
la_evaluate_nonlinear_models(fit)[, c("model_id", "RMSE", "R2")]

Extract model coefficients

Description

Returns a coefficient table from a fitted linear, nonlinear, or mixed model.

Usage

la_extract_coefficients(model)

Arguments

model

A fitted model object.

Value

A data.frame with coefficient names and estimates.

Display labels for leaf variables

Description

Returns user-friendly labels as either a named character vector or a plain character vector.

Usage

la_feature_display_names(named = FALSE)

Arguments

named

Logical. If TRUE, returns a named vector where names are the internal variable names and values are display labels.

Value

A character vector.

Display labels for leaf variables

Description

Returns a table of internal variable names and user-friendly labels used throughout the package interface.

Usage

la_feature_labels()

Value

A data.frame with internal names and display labels.

Fit candidate linear models for leaf area estimation

Description

Fits a default or user-supplied set of linear models using L, W, LA, and derived variables created automatically when needed.

Usage

la_fit_linear_models(
  data,
  formulas = NULL,
  include_no_intercept = TRUE,
  include_multiple = TRUE,
  include_polynomial = TRUE
)

Arguments

data

A data.frame containing at least L, W, and LA.
formulas

Optional named list of formulas. If NULL, the default formulas from la_linear_formulas() are used.
include_no_intercept

Logical. Only used when formulas = NULL.
include_multiple

Logical. Only used when formulas = NULL.
include_polynomial

Logical. Only used when formulas = NULL.

Value

A list with fitted models, formulas, modeling data, and a summary table enriched with equation information.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
names(fit$models)

Fit candidate linear mixed-effects models for leaf area estimation

Description

Fits a default or user-supplied set of linear mixed-effects models using L, W, LA, derived variables created automatically when needed, and a user-supplied grouping variable.

Usage

la_fit_mixed_models(
  data,
  group_var,
  formulas = NULL,
  random_slope = FALSE,
  include_multiple = TRUE,
  include_polynomial = TRUE,
  REML = FALSE,
  control = NULL
)

Arguments

data

A data.frame containing at least L, W, LA, and the grouping variable.
group_var

Character string with the grouping variable name.
formulas

Optional named list of formulas. If NULL, default formulas from la_mixed_formulas() are used.
random_slope

Logical. Only used when formulas = NULL.
include_multiple

Logical. Only used when formulas = NULL.
include_polynomial

Logical. Only used when formulas = NULL.
REML

Logical passed to lmer().
control

Optional control object passed to lmer().

Value

A list with fitted models, formulas, modeling data, and a summary table enriched with equation information.

Examples

data(leafarea_sample)
fit <- la_fit_mixed_models(leafarea_sample, group_var = "species")
names(fit$models)

Fit multiple nonlinear models

Description

Fits built-in or user-supplied nonlinear model specifications.

Usage

la_fit_nonlinear_models(
  data,
  models = NULL,
  specs = NULL,
  control = stats::nls.control(maxiter = 200, warnOnly = TRUE)
)

Arguments

data

A data.frame containing at least L, W, and LA.
models

Optional character vector of model IDs. Used only when specs is NULL.
specs

Optional named list of nonlinear model specifications.
control

Control list passed to stats::nls().

Value

A list containing fitted models, specifications, data, and summary.

Examples

data(leafarea_sample)
fit <- la_fit_nonlinear_models(leafarea_sample, models = c("power_LW", "exponential_LW"))
names(fit$models)
fit$summary[, c("model_id", "converged")]

Summarize a validated leaf area dataset

Description

Produces a concise overview of a validated dataset, including its dimensions, variable names, and summary statistics for L, W, and LA. This function is useful as a quick check before creating derived variables or fitting candidate models.

Usage

la_input_overview(data)

Arguments

data

A validated data.frame containing L, W, and LA.

Value

A list with the number of rows and columns, variable names, and a summary of the main measurement variables.

Examples

data(leafarea_sample)
validated_data <- la_validate_input(leafarea_sample)
overview <- la_input_overview(validated_data)
overview$n_rows
overview$summary

Extract fitted values from linear model results

Description

Extracts observed and fitted values for one selected linear model from the object returned by la_fit_linear_models().

Usage

la_linear_fitted_values(fit_object, model_id)

Arguments

fit_object

Object returned by la_fit_linear_models().
model_id

Character string with the model identifier.

Value

A data.frame with observed, fitted, and residual values.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
head(la_linear_fitted_values(fit, "lm_LW"))

List default linear model formulas

Description

Returns the default set of candidate linear formulas used by la_fit_linear_models().

Usage

la_linear_formulas(
  include_no_intercept = TRUE,
  include_multiple = TRUE,
  include_polynomial = TRUE
)

Arguments

include_no_intercept

Logical. If TRUE, includes the no-intercept model LA ~ 0 + LW.
include_multiple

Logical. If TRUE, includes multiple linear models.
include_polynomial

Logical. If TRUE, includes quadratic and cubic linear models based on derived variables.

Value

A named list of formulas.

Examples

names(la_linear_formulas())

List available derived variables

Description

Returns the default derived variables available in leafareaR.

Usage

la_list_derived()

Value

A character vector.

Examples

la_list_derived()

Calculate mean absolute error

Description

Calculate mean absolute error

Usage

la_mae(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_mae(obs, pred)

Calculate mean absolute percentage error

Description

Calculate mean absolute percentage error

Usage

la_mape(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value in percentage.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_mape(obs, pred)

Create a matrix plot for selected variables

Description

Generates a scatterplot matrix for selected numeric variables.

Usage

la_matrixplot(
  data,
  variables = NULL,
  hist = TRUE,
  pch = 19,
  cex = 0.6,
  col = "darkgreen",
  main = NULL
)

Arguments

data

A data.frame containing numeric variables.
variables

Character vector with variable names to include. If NULL, uses available default variables among L, W, LA.
hist

Logical; if TRUE, draws histograms on the diagonal.
pch

Plotting character for points.
cex

Point size.
col

Point color.
main

Optional main title.

Value

Invisibly returns the selected data used in the matrix plot.

Examples

data(leafarea_sample)

# To display the plot in an interactive R session, run:
# la_matrixplot(leafarea_sample, variables = c("L", "W", "LA"))

# CRAN-safe example:
grDevices::pdf(NULL)
la_matrixplot(leafarea_sample, variables = c("L", "W", "LA"))
grDevices::dev.off()

Create a default matrix plot for leaf variables

Description

Generates a matrix plot using available default variables among L, W, and LA.

Usage

la_matrixplot_default(
  data,
  hist = TRUE,
  pch = 19,
  cex = 0.6,
  col = "darkgreen",
  main = "Matrix plot of leaf variables"
)

Arguments

data

A data.frame containing leaf measurements.
hist

Logical; if TRUE, draws histograms on the diagonal.
pch

Plotting character for points.
cex

Point size.
col

Point color.
main

Optional main title.

Value

Invisibly returns the selected data used in the matrix plot.

Examples

data(leafarea_sample)

# To display the plot in an interactive R session, run:
# la_matrixplot_default(leafarea_sample)

# CRAN-safe example:
grDevices::pdf(NULL)
la_matrixplot_default(leafarea_sample)
grDevices::dev.off()

Calculate a standard metric table from observed and predicted values

Description

Calculate a standard metric table from observed and predicted values

Usage

la_metric_table(
  observed,
  predicted,
  n_parameters = NA_integer_,
  model_object = NULL,
  digits = 4
)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.
n_parameters

Optional number of estimated parameters.
model_object

Optional fitted model object used to extract AIC and BIC.
digits

Number of decimal places for rounding.

Value

A one-row data.frame.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_metric_table(obs, pred)

Extract coefficients from a mixed model

Description

Extract coefficients from a mixed model

Usage

la_mixed_coefficients(fit_object, model_id)

Arguments

fit_object

Object returned by la_fit_mixed_models().
model_id

Character string with the model identifier.

Value

A data.frame with fixed-effect coefficients.

Examples

data(leafarea_sample)
fit <- la_fit_mixed_models(leafarea_sample, group_var = "species")
la_mixed_coefficients(fit, names(fit$models)[1])

Extract fitted values from mixed-model results

Description

Extract fitted values from mixed-model results

Usage

la_mixed_fitted_values(fit_object, model_id)

Arguments

fit_object

Object returned by la_fit_mixed_models().
model_id

Character string with the model identifier.

Value

A data.frame with observed, fitted, residual, and group values.

Examples

data(leafarea_sample)
fit <- la_fit_mixed_models(leafarea_sample, group_var = "species")
head(la_mixed_fitted_values(fit, names(fit$models)[1]))

List default mixed-model formulas

Description

Returns the default set of candidate mixed-model formulas used by la_fit_mixed_models().

Usage

la_mixed_formulas(
  group_var,
  random_slope = FALSE,
  include_multiple = TRUE,
  include_polynomial = TRUE
)

Arguments

group_var

Character string with the grouping variable name.
random_slope

Logical. If TRUE, includes selected random-slope formulations.
include_multiple

Logical. If TRUE, includes multiple fixed-effect mixed models.
include_polynomial

Logical. If TRUE, includes quadratic and cubic mixed models based on derived variables.

Value

A named list of formulas.

Examples

names(la_mixed_formulas("species"))

Calculate mean squared error

Description

Calculate mean squared error

Usage

la_mse(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_mse(obs, pred)

Return coefficients from a selected nonlinear model

Description

Return coefficients from a selected nonlinear model

Usage

la_nonlinear_coefficients(fit_object, model_id)

Arguments

fit_object

Object returned by la_fit_nonlinear_models().
model_id

Character string with the model identifier.

Value

A data.frame with parameter estimates.

Examples

data(leafarea_sample)
fit <- la_fit_nonlinear_models(leafarea_sample, models = c("power_LW"))
la_nonlinear_coefficients(fit, "power_LW")

Extract observed, fitted values and residuals for a selected nonlinear model

Description

Extract observed, fitted values and residuals for a selected nonlinear model

Usage

la_nonlinear_fitted_values(fit_object, model_id)

Arguments

fit_object

Object returned by la_fit_nonlinear_models().
model_id

Character string with the model identifier.

Value

A data.frame with observed, fitted, and residual values.

Examples

data(leafarea_sample)
fit <- la_fit_nonlinear_models(leafarea_sample, models = c("power_LW"))
head(la_nonlinear_fitted_values(fit, "power_LW"))

Default nonlinear model specifications

Description

Returns the default built-in nonlinear candidate specifications.

Usage

la_nonlinear_specs()

Value

A named list of nonlinear model specifications.

Examples

names(la_nonlinear_specs())

Calculate Nash-Sutcliffe efficiency

Description

Calculate Nash-Sutcliffe efficiency

Usage

la_nse(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_nse(obs, pred)

Observed versus predicted leaf area plot

Description

Creates a plot comparing observed and predicted leaf area values.

Usage

la_plot_observed_predicted(
  observed,
  predicted,
  model_name = "Selected model",
  point_size = 2.2,
  alpha = 0.75
)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.
model_name

Character. Label used in the plot title.
point_size

Numeric. Point size.
alpha

Numeric. Point transparency.

Value

A ggplot object.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  fit <- la_fit_linear_models(leafarea_sample)
  vals <- la_linear_fitted_values(fit, model_id = "lm_LW")
  p <- la_plot_observed_predicted(vals$observed, vals$fitted, model_name = "lm_LW")
  print(p)
}

Histogram of residuals

Description

Creates a histogram of residuals from observed and predicted values.

Usage

la_plot_residual_histogram(
  observed,
  predicted,
  bins = 30,
  model_name = "Selected model"
)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.
bins

Number of histogram bins.
model_name

Character. Label used in the plot title.

Value

A ggplot object.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  fit <- la_fit_linear_models(leafarea_sample)
  vals <- la_linear_fitted_values(fit, model_id = "lm_LW")
  p <- la_plot_residual_histogram(vals$observed, vals$fitted, model_name = "lm_LW")
  print(p)
}

QQ plot of residuals

Description

Creates a QQ plot of residuals from observed and predicted values.

Usage

la_plot_residual_qq(observed, predicted, model_name = "Selected model")

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.
model_name

Character. Label used in the plot title.

Value

A ggplot object.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  fit <- la_fit_linear_models(leafarea_sample)
  vals <- la_linear_fitted_values(fit, model_id = "lm_LW")
  p <- la_plot_residual_qq(vals$observed, vals$fitted, model_name = "lm_LW")
  print(p)
}

Residuals versus fitted values plot

Description

Creates a residual diagnostic plot from observed and predicted values.

Usage

la_plot_residuals(
  observed,
  predicted,
  model_name = "Selected model",
  point_size = 2.2,
  alpha = 0.75
)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.
model_name

Character. Label used in the plot title.
point_size

Numeric. Point size.
alpha

Numeric. Point transparency.

Value

A ggplot object.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  fit <- la_fit_linear_models(leafarea_sample)
  vals <- la_linear_fitted_values(fit, model_id = "lm_LW")
  p <- la_plot_residuals(vals$observed, vals$fitted, model_name = "lm_LW")
  print(p)
}

Scatter plot between two selected variables

Description

Creates a scatter plot for two selected variables in a leaf area dataset.

Usage

la_plot_scatter(
  data,
  x,
  y = "LA",
  color_var = NULL,
  add_smooth = TRUE,
  point_size = 2.2,
  alpha = 0.75
)

Arguments

data

A data.frame containing the selected variables.
x

Character. Name of the x-axis variable.
y

Character. Name of the y-axis variable. Default is "LA".
color_var

Optional character. Grouping variable used for point color.
add_smooth

Logical. If TRUE, adds a linear trend line.
point_size

Numeric. Point size.
alpha

Numeric. Point transparency.

Value

A ggplot object.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  p <- la_plot_scatter(leafarea_sample, x = "L", y = "LA")
  print(p)
}

Scatter plots for multiple selected predictors against leaf area

Description

Creates a list of scatter plots using the selected predictors against the response variable.

Usage

la_plot_scatter_set(
  data,
  predictors = c("L", "W", "LW"),
  response = "LA",
  add_smooth = TRUE,
  point_size = 2.2,
  alpha = 0.75
)

Arguments

data

A data.frame containing the selected variables.
predictors

Character vector of predictor names.
response

Character. Name of the response variable.
add_smooth

Logical. If TRUE, adds a linear trend line.
point_size

Numeric. Point size.
alpha

Numeric. Point transparency.

Value

A named list of ggplot objects.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {
  data(leafarea_sample)
  dat <- la_create_derived(leafarea_sample, variables = c("LW"))
  plots <- la_plot_scatter_set(dat, predictors = c("L", "W", "LW"))
  print(plots[[1]])
}

Predict using one selected model from a fit object

Description

Predict using one selected model from a fit object

Usage

la_predict_from_results(
  fit_object,
  model_id = 1,
  newdata,
  allow_new_levels = TRUE,
  re_form = NULL
)

Arguments

fit_object

A fitted-model result object containing models.
model_id

Model position or model name.
newdata

A data.frame for prediction.
allow_new_levels

Logical used for mixed models.
re_form

Optional random-effects structure used for mixed models.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA")]
la_predict_from_results(fit, model_id = "lm_LW", newdata = newdata_ex)

Predict from a linear model

Description

Predict from a linear model

Usage

la_predict_linear_model(model, newdata)

Arguments

model

An object of class lm.
newdata

A data.frame for prediction.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA")]
la_predict_linear_model(fit$models[["lm_LW"]], newdata_ex)

Predict from a mixed model

Description

Predict from a mixed model

Usage

la_predict_mixed_model(model, newdata, allow_new_levels = TRUE, re_form = NULL)

Arguments

model

An object of class lmerMod.
newdata

A data.frame for prediction.
allow_new_levels

Logical used for mixed models.
re_form

Optional random-effects structure used for mixed models.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_mixed_models(leafarea_sample, group_var = "species")
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA", "species")]
la_predict_mixed_model(fit$models[[names(fit$models)[1]]], newdata_ex)

Predict from a fitted model

Description

Generic dispatcher for prediction from linear, nonlinear, or mixed models.

Usage

la_predict_model(
  model,
  newdata,
  model_type = c("auto", "linear", "nonlinear", "mixed"),
  allow_new_levels = TRUE,
  re_form = NULL
)

Arguments

model

A fitted model object.
newdata

A data.frame for prediction.
model_type

One of "auto", "linear", "nonlinear", or "mixed".
allow_new_levels

Logical used for mixed models.
re_form

Optional random-effects structure used for mixed models.

Value

A data.frame containing the prediction columns.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA")]
la_predict_model(fit$models[["lm_LW"]], newdata_ex)

Predict from a nonlinear model

Description

Predict from a nonlinear model

Usage

la_predict_nonlinear_model(model, newdata)

Arguments

model

An object of class nls.
newdata

A data.frame for prediction.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_nonlinear_models(leafarea_sample, models = c("power_LW"))
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA")]
la_predict_nonlinear_model(fit$models[["power_LW"]], newdata_ex)

Predict from the top-ranked model

Description

Predict from the top-ranked model

Usage

la_predict_top_ranked(
  ranked_table,
  fit_object,
  rank_position = 1,
  newdata,
  allow_new_levels = TRUE,
  re_form = NULL
)

Arguments

ranked_table

A ranked data.frame containing model_id.
fit_object

A fitted-model result object containing models.
rank_position

Row position within ranked_table.
newdata

A data.frame for prediction.
allow_new_levels

Logical used for mixed models.
re_form

Optional random-effects structure used for mixed models.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
ranked <- la_rank_models(met)
newdata_ex <- leafarea_sample[1:5, c("L", "W", "LA")]
la_predict_top_ranked(
  ranked,
  fit,
  rank_position = 1,
  newdata = newdata_ex
)

Calculate Pearson correlation coefficient

Description

Calculate Pearson correlation coefficient

Usage

la_r(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_r(obs, pred)

Calculate coefficient of determination

Description

Calculate coefficient of determination

Usage

la_r_squared(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_r_squared(obs, pred)

Rank models using a simple metric priority rule

Description

Orders models according to a sequence of evaluation criteria.

Usage

la_rank_models(
  metrics_table,
  sort_by = c("RMSE", "MAE", "CCC", "R2", "ABS_BIAS"),
  ascending = c(TRUE, TRUE, FALSE, FALSE, TRUE)
)

Arguments

metrics_table

A data.frame containing model metrics.
sort_by

Character vector with metric names used for ordering.
ascending

Logical vector indicating whether each metric should be sorted in ascending order.

Value

A ranked data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
la_rank_models(met)

Rank models by average metric positions

Description

Computes metric-wise ranks and aggregates them using the mean rank.

Usage

la_rank_models_by_metrics(metrics_table)

Arguments

metrics_table

A data.frame containing model metrics.

Value

A ranked data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
la_rank_models_by_metrics(met)

Rank models using a weighted score

Description

Computes a weighted score from selected metrics after min-max scaling.

Usage

la_rank_models_weighted(
  metrics_table,
  weights = list(RMSE = 0.3, MAE = 0.2, CCC = 0.2, R2 = 0.15, ABS_BIAS = 0.1, d = 0.05)
)

Arguments

metrics_table

A data.frame containing model metrics.
weights

Named list of metric weights.

Value

A ranked data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
la_rank_models_weighted(met)

Calculate root mean squared error

Description

Calculate root mean squared error

Usage

la_rmse(observed, predicted)

Arguments

observed

Numeric vector of observed values.
predicted

Numeric vector of predicted values.

Value

A numeric value.

Examples

obs <- c(10, 12, 15, 18)
pred <- c(9.8, 12.1, 14.7, 18.4)
la_rmse(obs, pred)

Select the top models from a ranking table

Description

Returns the first n rows from a ranked table.

Usage

la_top_models(
  ranking_table,
  n = 5,
  rank_column = c("rank_simple", "rank_weighted", "rank_mean")
)

Arguments

ranking_table

A ranked data.frame.
n

Number of rows to return.
rank_column

Column used to order the rows.

Value

A data.frame.

Examples

data(leafarea_sample)
fit <- la_fit_linear_models(leafarea_sample)
met <- la_evaluate_linear_models(fit)
ranked <- la_rank_models(met)
la_top_models(ranked, n = 3)

Validate and standardize input data for leaf area analysis

Description

Checks that the selected leaf length, leaf width, and observed leaf area columns are present and numeric, applies optional cleaning rules, and standardizes their names to L, W, and LA for downstream use in leafareaR.

Usage

la_validate_input(
  data,
  l_col = "L",
  w_col = "W",
  la_col = "LA",
  remove_na = TRUE,
  remove_nonpositive = TRUE,
  standardize_names = TRUE,
  keep_all_columns = FALSE
)

Arguments

data

A data.frame containing at least the columns L, W, and LA, or equivalent columns selected through l_col, w_col, and la_col.
l_col

Character. Name of the column containing leaf length.
w_col

Character. Name of the column containing leaf width.
la_col

Character. Name of the column containing observed leaf area.
remove_na

Logical. If TRUE, rows with missing values in the selected columns are removed.
remove_nonpositive

Logical. If TRUE, rows with values less than or equal to zero in the selected columns are removed.
standardize_names

Logical. If TRUE, the selected columns are renamed internally to L, W, and LA in the returned object.
keep_all_columns

Logical. If TRUE, keeps all original columns in the returned data.frame while standardizing the selected measurement columns.

Value

A validated data.frame ready for descriptive analysis, model fitting, prediction, and visualization in the leafareaR workflow.

Examples

data(leafarea_sample)
validated_data <- la_validate_input(leafarea_sample)
head(validated_data)

validated_with_groups <- la_validate_input(
  data = leafarea_sample,
  keep_all_columns = TRUE
)
head(validated_with_groups)

Example dataset for leaf area modeling

Description

A sample dataset included in leafareaR for testing data validation, descriptive statistics, derived variables, plotting, linear models, nonlinear models, mixed models, ranking, and prediction.

Usage

data(leafarea_sample)

Format

A data.frame with 9999 rows and 6 variables:

L

Leaf length.

W

Leaf width.

LA

Observed leaf area.

species

Species identifier.

block

Block identifier.

genotype

Genotype identifier.

Examples

data(leafarea_sample)
head(leafarea_sample)

Launch the built-in Shiny application

Description

Opens the interactive leafareaR Shiny app, which provides a graphical interface for loading example data or uploading user data, exploratory analysis, model fitting, evaluation, ranking, and prediction.

Usage

run_leafareaR_app(...)

Arguments

...

Additional arguments passed to shiny::runApp().

Value

Launches the application.

Examples

app_dir <- system.file("shiny", "leafareaR-app", package = "leafareaR")
dir.exists(app_dir)
if (interactive()) {
  run_leafareaR_app()
}