mbpls.data module

Module contents

The mbpls.data module contains methods to load data real world datasets and to create artificial data than can be used to test the mbpls methods.

mbpls.data.data_path()

mbpls.data.orthogonal_data(num_of_samples=11, params_block_one=4, params_block_two=4, params_block_three=4, num_of_variables_main_lin_comb=0, num_of_batches=1, random_state=None)

This function creates a dataset with three X-blocks, which are completely orthogonal amongst each other and one Y-block, that has two response variables, which are a linear combination of the variables defined for the three blocks.

num_of_samples: Amount of samples for the dataset params_block_one: Number of variables in the first block params_block_two: Number of variables in the second block params_block_three: Number of variables in the third block num_of_variables_main_lin_comb: Number of variables that are randon linear combinations of each variable (Multi-Colliniearity) num_of_batches: Number of batches for each block (third dimension)

X_1 = First X-block - Dimensionality ( num_of_samples, params_block_one*(num_of_variables_main_lin_comb+1), num_of_batches) X_2 = Second X-block - Dimensionality ( num_of_samples, params_block_two*(num_of_variables_main_lin_comb+1), num_of_batches) X_3 = Third X-block - Dimensionality ( num_of_samples, params_block_three*(num_of_variables_main_lin_comb+1), num_of_batches) Y = Y-block - Dimensionality (num_of_samples, 2, num_of_batches)

mbpls.data.load_CarbohydrateMicroarrays_Data()

mbpls.data.load_FTIR_Data()

mbpls.data.load_Intro_Data()