Preprocessing

v0.1.7

Data preparation that keeps column names: DataCleaner, Scaler, Encoder, Imputer, outlier detection/removal and train_test_split — every transformer returns DataFrames ready for models or export.

Quick Start

prepare.py

python

import synaptix as sx

df = sx.load_dataset("titanic")

cleaner = sx.preprocessing.DataCleaner()
cleaner.fit(df)
df_clean = cleaner.transform(df)      # same columns, clean values
df_new = cleaner.transform(new_data)  # apply the same cleaning later

scaler = sx.preprocessing.Scaler()
X = scaler.fit_transform(df_clean[["age", "fare"]])

encoder = sx.preprocessing.Encoder()
df_enc = encoder.fit_transform(df_clean)   # categorical -> numeric

API Reference

DataCleaner

DataCleaner()

pd.DataFrame

Scaler

Scaler(method='standard')

Encoder

Encoder()

Imputer

Imputer(strategy='mean')

detect_outliers / remove_outliers

detect_outliers(df, method='iqr') -> pd.Series / remove_outliers(df, method='iqr') -> pd.DataFrame

pd.Series (boolean mask) / pd.DataFrame

train_test_split

train_test_split(X, y, test_size=0.2, random_state=None)

tuple of DataFrames/arrays