experiment_impact_tracker package

Subpackages

• experiment_impact_tracker.cpu package
  • Submodules
  • experiment_impact_tracker.cpu.common module
  • experiment_impact_tracker.cpu.exceptions module
  • experiment_impact_tracker.cpu.intel module
  • experiment_impact_tracker.cpu.rapl module
  • Module contents
• experiment_impact_tracker.disk package
  • Submodules
  • experiment_impact_tracker.disk.common module
  • Module contents
• experiment_impact_tracker.emissions package
  • Submodules
  • experiment_impact_tracker.emissions.common module
  • experiment_impact_tracker.emissions.us_ca_parser module
  • Module contents
• experiment_impact_tracker.gpu package
  • Submodules
  • experiment_impact_tracker.gpu.exceptions module
  • experiment_impact_tracker.gpu.nvidia module
  • Module contents
• experiment_impact_tracker.py_environment package
  • Submodules
  • experiment_impact_tracker.py_environment.common module
  • Module contents

Submodules

experiment_impact_tracker.compute_tracker module

class experiment_impact_tracker.compute_tracker.ImpactTracker(logdir)

Bases: object

get_latest_info_and_check_for_errors()

launch_impact_monitor()

experiment_impact_tracker.compute_tracker.gather_initial_info(log_dir)

experiment_impact_tracker.compute_tracker.launch_power_monitor(queue, log_dir, initial_info, logger=None)

experiment_impact_tracker.compute_tracker.read_latest_stats(log_dir)

experiment_impact_tracker.constants module

experiment_impact_tracker.constants.load_regions_with_bounding_boxes()

Loads bounding boxes as shapely objects.

Returns

list of shapely objects containing regional geometries

Return type

list

experiment_impact_tracker.constants.read_terrible_json(path)

Reads a slightly malformed json file where each line is a different json dict.

Parameters

path (string) – the filepath to read from

Returns

list of dictionaries

Return type

[dict]

experiment_impact_tracker.create_graph_appendix module

experiment_impact_tracker.create_graph_appendix.create_graphs(input_path: str, output_path: str = '.', fig_x: int = 16, fig_y: int = 8, max_level=None)

experiment_impact_tracker.create_graph_appendix.create_scatterplot_from_df(df, x: str, y: str, output_path: str = '.', fig_x: int = 16, fig_y: int = 8)

Loads an executive summary df and creates a scatterplot from some pre-specified variables.

Parameters

• df ([type]) – [description]

• x (str) – [description]

• y (str) – [description]

• output_path (str, optional) – [description]. Defaults to ‘.’.

• fig_x (int, optional) – [description]. Defaults to 16.

• fig_y (int, optional) – [description]. Defaults to 8.

experiment_impact_tracker.create_graph_appendix.dateparse(time_in_secs)

experiment_impact_tracker.create_graph_appendix.handle_cpu_count_adjusted_average_load(df)

experiment_impact_tracker.data_info_and_router module

experiment_impact_tracker.data_utils module

experiment_impact_tracker.data_utils.load_data_into_frame(log_dir, max_level=None)

experiment_impact_tracker.data_utils.load_initial_info(log_dir)

experiment_impact_tracker.data_utils.log_final_info(log_dir)

experiment_impact_tracker.data_utils.safe_file_path(file_path)

experiment_impact_tracker.data_utils.write_csv_data_to_file(file_path, data, overwrite=False)

experiment_impact_tracker.data_utils.write_json_data_to_file(file_path, data, overwrite=False)

experiment_impact_tracker.data_utils.zip_data_and_info(log_dir, zip_path)

experiment_impact_tracker.data_utils.zip_files(src, dst, arcname=None)

Compress a list of files to a given zip

From https://stackoverflow.com/questions/16809328/zipfile-write-relative-path-of-files-reproduced-in-the-zip-archive

Parameters

• @src – Iterable object containing one or more element

• @dst – filename (path/filename if needed)

• @arcname – Iterable object containing the names we want to give to the elements in the archive (has to correspond to src)

experiment_impact_tracker.get_region_metrics module

experiment_impact_tracker.get_region_metrics.get_current_location()

experiment_impact_tracker.get_region_metrics.get_current_region_info()

experiment_impact_tracker.get_region_metrics.get_region_by_coords(coords)

experiment_impact_tracker.get_region_metrics.get_sorted_region_infos()

experiment_impact_tracker.get_region_metrics.get_zone_information_by_coords(coords)

experiment_impact_tracker.get_region_metrics.get_zone_name_by_id(zone_id)

experiment_impact_tracker.stats module

experiment_impact_tracker.stats.get_average_treatment_effect(data1, data2)

experiment_impact_tracker.stats.run_permutation_test(all_data, n1, n2)

experiment_impact_tracker.stats.run_test(test_id, data1, data2, alpha=0.05)

Compute tests comparing data1 and data2 with confidence level alpha

Taken from: https://arxiv.org/abs/1904.06979 Please cite that work if using this function.

Parameters

• test_id – (str) refers to what test should be used

• data1 – (np.ndarray) sample 1

• data2 – (np.ndarray) sample 2

• alpha – (float) confidence level of the test

Returns

(bool) if True, the null hypothesis is rejected

experiment_impact_tracker.utils module

experiment_impact_tracker.utils.gather_additional_info(info, logdir)

experiment_impact_tracker.utils.get_flop_count_tensorflow(graph=None, session=None)

experiment_impact_tracker.utils.get_timestamp(*args, **kwargs)

experiment_impact_tracker.utils.launch_power_monitorprocessify_func(q, *args, **kwargs)

experiment_impact_tracker.utils.processify(func)

Decorator to run a function as a process. Be sure that every argument and the return value is pickable. The created process is joined, so the code does not run in parallel.

experiment_impact_tracker.version module

Module contents