Data Transformation

Endpoint: POST /transform-data

Request Parameters

File Upload

• file (required): The file to be processed. Supported formats include CSV, Excel, etc.

Form Parameters

• method (required): The transformation method to apply. Supported values:
  • standardization: Standardizes numeric columns (mean = 0, standard deviation = 1).
  • minmax: Scales numeric columns to a specified range (default: [0, 1]).
  • log: Applies a log transformation to numeric columns.
  • power: Applies a power transformation (Yeo-Johnson) to numeric columns.

Processing Logic

1. Read the uploaded file: The file is converted into a Pandas DataFrame.
2. Validate the transformation method: Ensures the specified method is supported.
3. Identify numeric columns: Selects only numeric columns for transformation.
4. Handle missing values: Fills missing values in numeric columns with the mean.
5. Apply the transformation:
   • Standardization: Uses StandardScaler to standardize numeric columns.
   • Min-Max Scaling: Uses MinMaxScaler to scale numeric columns to the range [0, 1].
   • Log Transformation: Applies a log transformation (log1p) to numeric columns. If any values are ≤ 0, a shift is applied to ensure positivity.
   • Power Transformation: Applies a Yeo-Johnson power transformation using PowerTransformer. If any values are ≤ 0, a shift is applied to ensure positivity.
6. Combine transformed and non-numeric columns: Merges the transformed numeric columns with non-numeric columns.
7. Return the modified file: The transformed file is provided for download with the filename appended with _transformed_{method}.

Example Request

curl -X POST "http://localhost:8000/transform-data" \
-F "file=@data.csv" \
-F "method=standardization"