PREDICT

class predict(**kwargs)

Class containing all the functions from the PREDICT module.

Parameters

kwargsargument class

Specify any arguments from the PREDICT module (for a complete list of variables, visit the ROBERT documentation)

Parameters

destinationstr, default=None,

Directory to create the output file(s).

varfilestr, default=None

Option to parse the variables using a yaml file (specify the filename, i.e. varfile=FILE.yaml).

params_dirstr, default=''

Folder containing the database and parameters of the ML model.

csv_teststr, default=''

Name of the CSV file containing the test set (if any). A path can be provided (i.e. 'C:/Users/FOLDER/FILE.csv').

t_valuefloat, default=2

t-value that will be the threshold to identify outliers (check tables for t-values elsewhere). The higher the t-value the more restrictive the analysis will be (i.e. there will be more outliers with t-value=1 than with t-value = 4).

alphafloat, default=0.05

Significance level, or probability of making a wrong decision. This parameter is related to the confidence intervals (i.e. 1-alpha is the confidence interval). By default, an alpha value of 0.05 is used, which corresponds to a confidence interval of 95%.

shap_showint, default=10,

Number of descriptors shown in the plot of the SHAP analysis.

pfi_showint, default=10,

Number of descriptors shown in the plot of the PFI analysis.

pfi_epochsint, default=5,

Sets the number of times a feature is randomly shuffled during the PFI analysis (standard from sklearn webpage: 5).

namesstr, default=''

Column of the names for each datapoint. Names are used to print outliers.