Skip to main content

PanDas PRePRocessor: Preprocess Pandas Objects for Machine Learning

Project description

PanDas PRePRocessor: Preprocess Pandas Objects for Machine Learning


$ pip install pdprpr


Assume you have following DataFrame to be preprocessed:

from pandas import DataFrame

df = DataFrame({
    'num': [1, 3, float('nan')],  # numerical feature, needs to be scaled in [0, 1]
    'cat': ['p', 'q', 'r'],       # categorical feature, needs to be transformted to dummy var
    'bin': [False, False, True],  # binary feature, needs to be 0 / 1
}, columns =['num', 'cat', 'bin'])
#    num cat    bin
# 0  1.0   p  False
# 1  3.0   q  False
# 2  NaN   r   True

Define preprocessing settings:

# preprocessing.yml
- name: num
  kind: numerical

- name: cat
  kind: categorical

- name: bin
  kind: binary

Then create DataFramePreprocessor with them:

import yaml

with open('preprocessing.yml') as f:
    settings = yaml.load(f)

from pdprpr import DataFramePreprocessor

processor = DataFramePreprocessor(settings)

Finally use it to preprocess the DataFrame:

#    num__VALUE  cat__p  cat__q  cat__r  bin__TRUE
# 0         0.0       1       0       0          0
# 1         1.0       0       1       0          0
# 2         NaN       0       0       1          1


For more options please see tests untill docs get available…

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pdprpr-0.7.2.tar.gz (4.4 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page