dataliner 1.1.2

A dataprocessing package for data preprocess and feature engineering.

DataLiner - Data processing package for Python

A dataprocessing package for data preprocess and feature engineering.
Please feel free to send pull requests for bug fix, improvements or new preprocessing methods!

Installation

! pip install dataliner

Quick Start

Train data from Kaggle Titanic is used in this example. https://www.kaggle.com/c/titanic/data

import pandas as pd
from sklearn.pipeline import make_pipeline
import dataliner as dl

df = pd.read_csv('train.csv')
target_col = 'Survived'
X = df.drop(target_col, axis=1)
y = df[target_col]

process = make_pipeline(
    dl.DropNoVariance(),
    dl.DropHighCardinality(),
    dl.BinarizeNaN(),
    dl.ImputeNaN(),
    dl.TargetMeanEncoding(),
    dl.DropHighCorrelation(),
    dl.StandardizeData(),
    dl.DropLowAUC(),
)

process.fit_transform(X, y)

Documentation

https://shallowdf20.github.io/dataliner/preprocessing.html

What to expect in next version (1.1):

• Specifying columns to apply each preprocess
• Accepting numpy data types (currently pandas DataFrame / Series only)
• Additional preprocess (Data Normalization, Frequency Encoding, etc.)
• Strictly keeping the order of column names after preprocess