sharkpy.reporting

Functions

_create_temp_plot(→ Optional[str])

Create a temporary plot and return its path

_get_feature_importance_section(→ Tuple[List[str], ...)

Generate feature importance section for the report and a DataFrame for DOCX table

_get_statistical_details_section(→ List[str])

Generate statistical details section for the report

_add_table_to_doc(doc, df, title)

Helper function to add a pandas DataFrame as a table to DOCX

_export_docx_report(path, shark, cv_metrics_df, ...)

Export report as Word document with enhanced formatting, tables, and plots.

_export_txt_report(path, lines)

Export report as text file

_convert_docx_to_pdf(→ None)

Convert DOCX to PDF using available tools

report(→ Tuple[Dict[str, numpy.ndarray], Dict[str, float]])

Generate comprehensive model performance report including cross-validation.

Module Contents

sharkpy.reporting._create_temp_plot(shark: Any, kind: str, width: int = 8, height: int = 6) → str | None[source]

Create a temporary plot and return its path

sharkpy.reporting._get_feature_importance_section(shark: Any) → Tuple[List[str], pandas.DataFrame | None][source]

Generate feature importance section for the report and a DataFrame for DOCX table

sharkpy.reporting._get_statistical_details_section(shark: Any) → List[str][source]

Generate statistical details section for the report

sharkpy.reporting._add_table_to_doc(doc: docx.Document, df: pandas.DataFrame, title: str)[source]

Helper function to add a pandas DataFrame as a table to DOCX

sharkpy.reporting._export_docx_report(path: str, shark: Any, cv_metrics_df: pandas.DataFrame, train_metrics_df: pandas.DataFrame, problem_type: str)[source]

Export report as Word document with enhanced formatting, tables, and plots.

sharkpy.reporting._export_txt_report(path: str, lines: List[str])[source]

Export report as text file

sharkpy.reporting._convert_docx_to_pdf(docx_path: str, pdf_path: str) → None[source]

Convert DOCX to PDF using available tools

sharkpy.reporting.report(self, cv_folds: int = 5, export_path: str | None = None, format: str = 'txt') → Tuple[Dict[str, numpy.ndarray], Dict[str, float]][source]

Generate comprehensive model performance report including cross-validation.

Parameters:

  • self (Any) – The Shark instance

  • cv_folds (int, optional) – Number of folds for K-Fold cross-validation (default: 5)

  • export_path (str, optional) – Path to export the report. If None, report is only printed. If a directory is provided, a timestamped file will be created.

  • format (str, optional) – Export format: ‘txt’, ‘pdf’, or ‘docx’ (default: ‘txt’)

Returns:

  • cv_resultsdict

    Dictionary containing cross-validation results

  • train_metricsdict

    Dictionary containing training set metrics

Return type:

tuple

Notes

  • For PDF export, ensure Microsoft Word or LibreOffice is installed for docx2pdf conversion.

  • Visualizations include feature importance, predictions/residuals (regression), or confusion matrix/ROC/PR curves (classification).