DataReporting

Submodules

reporting module

hardy.data_reporting.reporting.model_analysis(model, test_set, test_set_list=None)

The function that provides analysis of a trained model for its predicted output and actual output

Parameters:

model: keras.model

a keras instance of the trained model to use for the prediction

test_set_list: list

list representing the file names, images and labels for test set.

test_set: keras.ImageDataGenerator iterator

the interator instance created using the keras.ImageDataGenerator. This can either be a NumpyArrayIterator or a DirectoryIterator

Returns:

result: pandas.DataFrame

dataframe having filenames, actual labels, predicted labels and probability for decision

hardy.data_reporting.reporting.report_dataframes(report_path)

hardy.data_reporting.reporting.report_plots(hyperparam_df, history_df)

hardy.data_reporting.reporting.summary_dataframe(report_path)

hardy.data_reporting.reporting.summary_report_plots(report_path)

The function that plots the parallel coordinates between report name, layers, optimizer, activation function, kernel size, pooling and accuracy.

Parameters:

report_path: str

string representing the location of parent report directory

hardy.data_reporting.reporting.summary_report_tables(report_path)

The function that returns tables wiht the summary of the transformations used and they performnce

Parameters:

report_path: str

string representing the location of parent report directory

Returns:

summary_df : pandas Dataframe

Table containing information of the transformations run, which data series they were applied to and their plot format

tform_rank_df : pandas Dataframe

Table containing information of the run anme and its overall performance

Module contents