AnalysisWeb 0.0.3

AnalysisWeb

AnalysisWeb Package

you can install this package by

pip install AnalysisWeb

The package consists of 5 function :

• 1 create_results_index: Creates an HTML index page linking to all result files.
• 2 config_to_html : Creates an HTML report of the configuration object with compact layout.
• 3 image_report_to_html : Creates an HTML report with a base64 image and dictionary information.
• 4 save_table_html : Saves a DataFrame as an HTML file using template markers for clean appending
• 5 image_gallery_to_html : Create an HTML page with multiple base64 images in a gallery layout.

Use the functions above to create the *.html pages for each run

Sequencer

The package consists of the Sequencer Class which helps maintain the CSV file system required for the index page

For each process (eg. "fit", "valid", "analysis" ...) set the result as a string (eg. "Fitting") use the Sequencer to connect between the processes and update the status.

Example usage

sequencer = Sequencer(
    plots_data_file, model.timestamp, config.model_type)


sequencer.start()

sequencer.add_algorithm(
    model.analyze, holdout_sets=holdout_sets, training_sets=train_sets
)

sequencer.add_algorithm(model.compute_yield, holdout_sets=holdout_sets)

sequencer.add_algorithm(
    model.validation,
    test_sets=holdout_sets,
    plot_label=f"holdout_{args.model_type}",
)

score = model.scoring()
sequencer.add_score("score", score)

sequencer.end()

Web APP

you can launch the webapp with

python -m AnalysisWeb.app --results-dir --csv-dir