methodDR: Generic functions for deepregression models
In davidruegamer/deepregression: Fitting Deep Distributional Regression

plot.deepregression

R Documentation

Generic functions for deepregression models

Description

Generic functions for deepregression models

Function to prepare data for deepregression use

Predict based on a deepregression object

Predict based on a deeptrafo object

Function to extract fitted distribution

Fit a deepregression model

Fit a custom deepregression models

Print function for deepregression model

Cross-validation for deepgression objects

mean of model fit

Standard deviation of fit distribution

Calculate the distribution quantiles

Usage

## S3 method for class 'deepregression'
plot(
  x,
  which = NULL,
  which_param = 1,
  plot = TRUE,
  use_posterior = FALSE,
  grid_length = 40,
  ...
)

prepare_data(x, data, pred = TRUE)

## S3 method for class 'deepregression'
predict(
  object,
  newdata = NULL,
  batch_size = NULL,
  apply_fun = tfd_mean,
  convert_fun = as.matrix,
  dtype = "float32",
  ...
)

## S3 method for class 'deeptrafo'
predict(object, newdata = NULL, which = NULL, cast_float = FALSE, ...)

## S3 method for class 'deepregression'
fitted(object, apply_fun = tfd_mean, ...)

## S3 method for class 'deepregression'
fit(
  x,
  early_stopping = FALSE,
  verbose = TRUE,
  view_metrics = FALSE,
  patience = 20,
  save_weights = FALSE,
  auc_callback = FALSE,
  validation_data = NULL,
  callbacks = list(),
  convertfun = function(x) tf$constant(x, dtype = "float32"),
  ...
)

## S3 method for class 'deepregression'
coef(object, variational = FALSE, params = NULL, type = NULL, ...)

## S3 method for class 'deepregression'
print(x, ...)

cv(
  x,
  verbose = FALSE,
  patience = 20,
  plot = TRUE,
  print_folds = TRUE,
  cv_folds = NULL,
  stop_if_nan = TRUE,
  mylapply = lapply,
  save_weights = FALSE,
  callbacks = list(),
  ...
)

## S3 method for class 'deepregression'
mean(x, data = NULL, ...)

## S3 method for class 'deepregression'
sd(x, data = NULL, ...)

## S3 method for class 'deepregression'
quantile(x, data = NULL, probs, ...)

Arguments

`x`	a deepregression object
`which`	which effect to plot, default selects all.
`which_param`	integer of length 1. Corresponds to the distribution parameter for which the effects should be plotted.
`plot`	whether to plot the resulting losses in each fold
`use_posterior`	logical; if `TRUE` it is assumed that the strucuted_nonlinear layer has stored a list of length two as weights, where the first entry is a vector of mean and sd for each network weight. The sd is transformed using the `exp` function. The plot then shows the mean curve +- 2 times sd.
`grid_length`	the length of an equidistant grid at which a two-dimensional function is evaluated for plotting.
`...`	arguments passed to the `predict` function
`data`	either `NULL` or a new data set
`pred`	logical, where the data corresponds to a prediction task
`object`	a deepregression model
`newdata`	optional new data, either data.frame or list
`batch_size`	batch_size for generator (image use cases)
`apply_fun`	function applied to fitted distribution, per default `tfd_mean`
`convert_fun`	how should the resulting tensor be converted, per default `as.matrix`
`dtype`	string for conversion
`early_stopping`	logical, whether early stopping should be user.
`verbose`	whether to print training in each fold
`view_metrics`	logical, whether to trigger the Viewer in RStudio / Browser.
`patience`	number of patience for early stopping
`save_weights`	logical, whether to save weights in each epoch.
`auc_callback`	logical, whether to use a callback for AUC
`validation_data`	optional specified validation data
`callbacks`	a list of callbacks used for fitting
`convertfun`	function to convert R into Tensor object
`variational`	logical, if TRUE, the function takes into account that coefficients have both a mean and a variance
`params`	integer, indicating for which distribution parameter coefficients should be returned (default is all parameters)
`type`	either NULL (all types of coefficients are returned), "linear" for linear coefficients or "smooth" for coefficients of smooth terms
`print_folds`	whether to print the current fold
`cv_folds`	see `deepregression`
`stop_if_nan`	logical; whether to stop CV if NaN values occur
`mylapply`	lapply function to be used; defaults to `lapply`
`probs`	the quantile value(s)
`train`	function taking the keras model, inputs and outputs as an updates the model
`epochs`	integer; the number of epochs to train
`print_fun`	function to print metrics