predict_keras: Predict with keras
In jmaspons/MLTools: Pipes and tools for machine learning

predict_keras

R Documentation

Predict with keras

Description

Predict with keras

Usage

predict_keras(
  modelNN,
  predInput,
  maskNA = NULL,
  scaleInput = FALSE,
  col_means_train,
  col_stddevs_train,
  batch_size = NULL,
  filename = "",
  tempdirRaster = NULL,
  nCoresRaster = 2
)

Arguments

`modelNN`	an `keras::keras()` model.
`predInput`	`data.frame` or `raster` with colnames or layer names matching the expected input for modelRF.
`maskNA`	value to assign to `NA`s after scaling and passed to `keras::layer_masking()`.
`scaleInput`	if `TRUE`, scale `predInput` with `col_means_train` and col `col_stddevs_train`.
`col_means_train`	the original mean of the `predInput` columns.
`col_stddevs_train`	the original sd of the `predInput` columns.
`batch_size`	for fit and predict functions. The bigger the better if it fits your available memory. Integer or "all".
`filename`	the file to write the raster predictions.
`tempdirRaster`	path to a directory to save temporal raster files.
`nCoresRaster`	number of cores used for parallelized raster cores. Use half of the available cores by default.