Local Feature Preprocess

Introduction

Local feature preprocess is a module that preprocess feature locally.

Parameter List

identity: str Federated identity of the party, should be one of label_trainer or trainer.

model_info:

  • name: str Model name, should be local_feature_preprocess.

input:
  • trainset:

    • type: str Train dataset type, support csv.

    • path: str If type is csv, folder path of train dataset.

    • name: str If type is csv, file name of train dataset.

    • has_id: bool If type is csv, whether dataset has id column.

    • has_label: bool If type is csv, whether dataset has label column.

  • valset:

    • type: str Validation dataset type, support csv.

    • path: str If type is csv, folder path of validation dataset.

    • name: str If type is csv, file name of validation dataset.

    • has_id: bool If type is csv, whether dataset has id column.

    • has_label: bool If type is csv, whether dataset has label column.

output:

  • path: str Folder path of output model.

  • model:

    • name: str File name of output model.

  • trainset:

    • name: str File name of output trainset.

  • valset:

    • name: str File name of output valset.

train_info:
  • train_params:
    • missing:

      • missing_values: int or float or str or list The placeholder for the missing values.

      • strategy: str The imputation strategy.

      • fill_value: str or numerical value When strategy == “constant”, fill_value is used to replace all occurrences of missing_values.

      • missing_features: map Fine-grained configuration for column preprocess(axis = 0). The format is defined as: {column name: {“missing_values”: placeholder for the missing values, “strategy”: imputation strategy, “fill_value”: imputation value when strategy == “constant”}, …}.

    • outlier:

      • outlier_values: int or float or str or list The placeholder for the outlier values.

      • outlier_features: map Fine-grained configuration for column preprocess(axis = 0). The format is defined as: {column name: {“outlier_values”: placeholder for the outlier values}, …}.

    • onehot:

      • onehot_features: map Fine-grained configuration for column preprocess(axis = 0). The format is defined as: {column name: {}, …, column name: {}}.