CLI first commit (drevalpy --help and drevalpy-report --help

Author: PascalIversen

create_report.py

@@ -0,0 +1,13 @@
+"""Main script to run the drug response evaluation pipeline."""
+
+from drevalpy.utils import get_parser, main
+
+
+def cli_main():
+    """Command line interface entry point for the drug response evaluation pipeline."""
+    args = get_parser().parse_args()
+    main(args)
+
+
+if __name__ == "__main__":
+    cli_main()

drevalpy/cli.py

@@ -0,0 +1,208 @@
+"""Generate evaluation reports after running a drug response experiment."""
+
+import argparse
+import os
+import pathlib
+from collections.abc import Iterable
+from typing import Union
+
+import pandas as pd
+
+from drevalpy.visualization.utils import (
+    create_html,
+    create_index_html,
+    create_output_directories,
+    draw_algorithm_plots,
+    draw_test_mode_plots,
+    parse_results,
+    prep_results,
+    write_results,
+)
+
+
+def generate_reports_for_test_mode(
+    test_mode: str,
+    evaluation_results: pd.DataFrame,
+    evaluation_results_per_drug: pd.DataFrame,
+    evaluation_results_per_cell_line: pd.DataFrame,
+    true_vs_pred: pd.DataFrame,
+    run_id: str,
+    path_data: Union[str, pathlib.Path],
+    result_path: Union[str, pathlib.Path],
+) -> None:
+    """
+    Generate reports (plots and HTML) for a single test mode.
+
+    :param test_mode: The test mode to generate reports for.
+    :param evaluation_results: Aggregated evaluation results.
+    :param evaluation_results_per_drug: Evaluation results per drug.
+    :param evaluation_results_per_cell_line: Evaluation results per cell line.
+    :param true_vs_pred: True vs predicted values.
+    :param run_id: Unique run identifier.
+    :param path_data: Path to the dataset directory.
+    :param result_path: Path to the results directory.
+    """
+    print(f"Generating report for {test_mode} ...")
+    unique_algos: list[str] = draw_test_mode_plots(
+        test_mode=test_mode,
+        ev_res=evaluation_results,
+        ev_res_per_drug=evaluation_results_per_drug,
+        ev_res_per_cell_line=evaluation_results_per_cell_line,
+        custom_id=run_id,
+        path_data=path_data,
+        result_path=result_path,
+    )
+    unique_algos_set = set(unique_algos) - {
+        "NaiveMeanEffectsPredictor",
+        "NaivePredictor",
+        "NaiveCellLineMeansPredictor",
+        "NaiveTissueMeansPredictor",
+        "NaiveDrugMeanPredictor",
+    }
+    for algorithm in unique_algos_set:
+        draw_algorithm_plots(
+            model=algorithm,
+            ev_res=evaluation_results,
+            ev_res_per_drug=evaluation_results_per_drug,
+            ev_res_per_cell_line=evaluation_results_per_cell_line,
+            t_vs_p=true_vs_pred,
+            test_mode=test_mode,
+            custom_id=run_id,
+            result_path=result_path,
+        )
+
+    all_files: list[str] = []
+    for _, _, files in os.walk(f"{result_path}/{run_id}"):
+        for file in files:
+            if file.endswith("json") or (
+                file.endswith(".html") and file not in ["index.html", "LPO.html", "LCO.html", "LDO.html"]
+            ):
+                all_files.append(file)
+
+    create_html(
+        run_id=run_id,
+        test_mode=test_mode,
+        files=all_files,
+        prefix_results=f"{result_path}/{run_id}",
+    )
+
+
+def generate_reports_for_all_test_modes(
+    test_modes: Iterable[str],
+    evaluation_results: pd.DataFrame,
+    evaluation_results_per_drug: pd.DataFrame,
+    evaluation_results_per_cell_line: pd.DataFrame,
+    true_vs_pred: pd.DataFrame,
+    run_id: str,
+    path_data: Union[str, pathlib.Path],
+    result_path: Union[str, pathlib.Path],
+) -> None:
+    """
+    Generate reports for all test modes.
+
+    :param test_modes: Iterable of test modes to process.
+    :param evaluation_results: Aggregated evaluation results.
+    :param evaluation_results_per_drug: Evaluation results per drug.
+    :param evaluation_results_per_cell_line: Evaluation results per cell line.
+    :param true_vs_pred: True vs predicted values.
+    :param run_id: Unique run identifier.
+    :param path_data: Path to the dataset directory.
+    :param result_path: Path to the results directory.
+    """
+    for test_mode in test_modes:
+        generate_reports_for_test_mode(
+            test_mode=test_mode,
+            evaluation_results=evaluation_results,
+            evaluation_results_per_drug=evaluation_results_per_drug,
+            evaluation_results_per_cell_line=evaluation_results_per_cell_line,
+            true_vs_pred=true_vs_pred,
+            run_id=run_id,
+            path_data=path_data,
+            result_path=result_path,
+        )
+
+
+def render_report(
+    run_id: str,
+    dataset: str,
+    path_data: Union[str, pathlib.Path] = "data",
+    result_path: Union[str, pathlib.Path] = "results",
+) -> None:
+    """
+    Render a full evaluation report pipeline.
+
+    :param run_id: Unique run identifier for locating results.
+    :param dataset: Dataset name to filter results.
+    :param path_data: Path to the dataset directory. Defaults to "data".
+    :param result_path: Path to the results directory. Defaults to "results".
+
+    :raises AssertionError: If the folder with the run_id does not exist under result_path.
+    """
+    path_data = pathlib.Path(path_data).resolve()
+    result_path = pathlib.Path(result_path).resolve()
+
+    if not os.path.exists(f"{result_path}/{run_id}"):
+        raise AssertionError(f"Folder {result_path}/{run_id} does not exist. The pipeline has to be run first.")
+
+    (
+        evaluation_results,
+        evaluation_results_per_drug,
+        evaluation_results_per_cell_line,
+        true_vs_pred,
+    ) = parse_results(path_to_results=f"{result_path}/{run_id}", dataset=dataset)
+
+    (
+        evaluation_results,
+        evaluation_results_per_drug,
+        evaluation_results_per_cell_line,
+        true_vs_pred,
+    ) = prep_results(
+        evaluation_results, evaluation_results_per_drug, evaluation_results_per_cell_line, true_vs_pred, path_data
+    )
+
+    write_results(
+        path_out=f"{result_path}/{run_id}/",
+        eval_results=evaluation_results,
+        eval_results_per_drug=evaluation_results_per_drug,
+        eval_results_per_cl=evaluation_results_per_cell_line,
+        t_vs_p=true_vs_pred,
+    )
+
+    create_output_directories(result_path, run_id)
+    test_modes = evaluation_results["test_mode"].unique()
+
+    generate_reports_for_all_test_modes(
+        test_modes=test_modes,
+        evaluation_results=evaluation_results,
+        evaluation_results_per_drug=evaluation_results_per_drug,
+        evaluation_results_per_cell_line=evaluation_results_per_cell_line,
+        true_vs_pred=true_vs_pred,
+        run_id=run_id,
+        path_data=path_data,
+        result_path=result_path,
+    )
+
+    create_index_html(
+        custom_id=run_id,
+        test_modes=test_modes,
+        prefix_results=f"{result_path}/{run_id}",
+    )
+
+
+def main() -> None:
+    """
+    Command line interface entry point for rendering evaluation reports.
+
+    Parses command line arguments and calls render_report.
+    """
+    parser = argparse.ArgumentParser(description="Generate reports from evaluation results")
+    parser.add_argument("--run_id", required=True, help="Run ID for the current execution")
+    parser.add_argument("--dataset", required=True, help="Dataset name for which to render the result file")
+    parser.add_argument("--path_data", default="data", help="Path to the data")
+    parser.add_argument("--result_path", default="results", help="Path to the results")
+    args = parser.parse_args()
+    render_report(args.run_id, args.dataset, args.path_data, args.result_path)
+
+
+if __name__ == "__main__":
+    main()

drevalpy/visualization/create_report.py

@@ -6,6 +6,10 @@ authors = ["DrEvalPy development team"]
 license = "GPL-3.0"
 readme = "README.md"
 
+[tool.poetry.scripts]
+drevalpy = "drevalpy.cli:cli_main"
+drevalpy-report = "drevalpy.visualization.create_report:main"
+
 [tool.poetry.dependencies]
 python = ">=3.11,<3.14"
 numpy = ">=1.20"