Data Management Skill

Overview

This skill provides comprehensive data management capabilities including loading data from Excel/CSV files, filtering DataFrames by column values, applying transformations, managing data arrays, and building data pipelines. All operations are configurable via YAML files for reproducible data workflows.

Key Components

DataManagement Class (data_management.py)

High-level data pipeline management:

• router(cfg)

  - Route data operations based on configuration

• get_df_data(cfg)

  - Load DataFrame from configuration

• get_df_array_from_cfg(cfg)

  - Load multiple DataFrames as array

• get_filtered_df(data_set_cfg, df)

  - Apply filters to DataFrame

• get_transformed_df(data_set_cfg, df)

  - Apply transformations to DataFrame

ReadFromExcel Class (data.py)

Excel file reading with sheet selection:

• from_xlsx(cfg, file_index)

  - Read Excel files with configurable sheet selection
• Supports multiple sheets, header row configuration, data range selection

ReadFromCSV Class (data.py)

CSV file reading with encoding detection:

• to_df(cfg, file_index)

  - Read CSV to DataFrame
• Automatic encoding detection with chardet
• Configurable delimiter, header options

ReadData Class (data.py)

Advanced data reading operations:

• df_filter_by_column_values(cfg, df, file_index)

  - Filter DataFrame by column values

• xlsx_to_df_by_keyword_search(cfg)

  - Read Excel by keyword-based row search

• get_data_from_xlsx_and_csv(cfg)

  - Unified Excel/CSV reading

Usage Patterns

Data Loading Configuration

data:
  files:
    - path: "data.xlsx"
      sheet_name: "Sheet1"
      header_row: 0
      columns: ["A", "B", "C"]

Filtering Configuration

data:
  filter:
    column: "status"
    values: ["active", "pending"]
    operator: "in"  # in, equals, gt, lt, contains

Transformation Configuration

data:
  transform:
    - type: "rename"
      mapping: {"old_col": "new_col"}
    - type: "add_column"
      name: "calculated"
      expression: "col_a + col_b"

Common Workflows

1. Excel Pipeline: Load Excel → Filter rows → Transform columns → Export
2. Multi-Source Merge: Load CSV + Excel → Merge on key → Validate → Save
3. Data Validation: Load data → Apply filters → Check constraints → Report
4. Batch Processing: Config with file list → Process each → Aggregate results

Module Location

• Pipeline:

  src/assetutilities/common/data_management.py

• Readers:

  src/assetutilities/common/data.py

Dependencies

• pandas (DataFrame operations)
• openpyxl (Excel reading)
• chardet (encoding detection)
• numpy (numerical operations)