Configuration files defining preprocessing parameters.

The different configuration files are used in the models with the respective name. See irp_preprocess().

Usage

model_klason_lignin_2_config

model_holocellulose_2_config

model_eac_1_config

model_edc_1_config

A list with the following elements:

irp_preprocess

A list with the a subset of the following elements:

do_interpolate: See irp_preprocess().
interpolate_start: See irp_preprocess().
interpolate_dw: See irp_preprocess().
do_clip: See irp_preprocess().
clip_range: See irp_preprocess().
do_interpolate_region: See irp_preprocess().
interpolate_region_range: See irp_preprocess().
do_bc: See irp_preprocess().
bc_method: See irp_preprocess().
bc_degree: See irp_preprocess().
bc_cutoff: See irp_preprocess().
bc_do_impute: See irp_preprocess().
do_smooth: See irp_preprocess().
smooth_method: See irp_preprocess().
smooth_p: See irp_preprocess().
smooth_n: See irp_preprocess().
smooth_m: See irp_preprocess().
smooth_ts: See irp_preprocess().
smooth_k: See irp_preprocess().
do_normalise: See irp_preprocess().
normalise_method: See irp_preprocess().
do_bin: See irp_preprocess().
bin_width: See irp_preprocess().
do_scale: See irp_preprocess().
scale_center: See irp_preprocess(). This is used to preprocess the training data before model computation. The values in data_scale are used to scale data used for prediction.
scale_scale: See irp_preprocess(). This is used to preprocess the training data before model computation. The values in data_scale are used to scale data used for prediction.

data_scale

Values used for scaling data used for prediction. A list with the following elements:

y_center: A numeric value representing the value to subtract from the dependent variable of a model during scaling. Corresponds to argument center in base::scale().
y_scale: A numeric value representing the value by which the dependent variable of a model is divided during scaling. Corresponds to argument scale in base::scale().
x_center: The same as y_center, but for the independent variable(s). Can be a numeric vector with an element for each independent variable.
y_scale: The same as y_center, but for the independent variable(s). Can be a numeric vector with an element for each independent variable.

An object of class list of length 2.