Skip to main content

Automated view of dataset

Project description

Description

Requirements

  • python >= 3.6
  • numpy >= 1.16
  • pandas >= 0.23
  • matplotlib >= 3.02
  • IPython >= 7.2.0

Module, provides the function view, which displays general information on the data:

  • Result of method info()
  • Result of method describe()
  • for numeric / categorical variables - The number of nulls in the data (amount and percentage for each column)
  • grid of histograms for numeric variables
  • Top-5 of the most frequent categorical variables (for each)
  • Split count statistic by category with less 5 unique values
  • matrix correlation
  • dataframe of higher correlation pair (over 0.7)

Parameters (function view):

  • d - table with data
  • only_numeric - True / False, default: True. True - information output only by numerical variables, False - information output by numerical and categorical variables.
  • full_stats - True / False, default: False. False - output information on numerical variables without interquartile range, data boundaries without outliers, True - complete output with IQR, min and max by IQR and amount of outliers.
  • histograms - True / False, default: True. True - output with building histograms for numerical variables, False - without building histograms

Top-5 elements of categorical variables

The postfix (_name / _count) is assigned to the name of the data column to create top-5 elements table:

  • _name - category name
  • _count - number of unique elements in this category. If there are less than 5 unique elements, then the values ​​in the _count field are filled -1

Count statistic of numbers variables split by some categorical variables

  • if number of unique elements in category variable less 5, then show count statistic

Correlation matrix

  • Correlation matrix with heatmap (pearson corr index)

Install

$ pip install data_view

Usage

$ python3

from data_view import *

d = pd.DataFrame(np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]]), columns=['a', 'b', 'c'])

view(d, only_numeric=True, histograms=False)

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

data-view-0.2.0.tar.gz (5.4 kB view details)

Uploaded Source

File details

Details for the file data-view-0.2.0.tar.gz.

File metadata

  • Download URL: data-view-0.2.0.tar.gz
  • Upload date:
  • Size: 5.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/2.0.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.6.0 requests-toolbelt/0.9.1 tqdm/4.36.1 CPython/3.7.5

File hashes

Hashes for data-view-0.2.0.tar.gz
Algorithm Hash digest
SHA256 8871e19f0db0d37ece18933c5d40476f53c863b8762c251c10f69cdfa86ad6ed
MD5 d049527b86aef81f8f98c367697c3599
BLAKE2b-256 583f61747f454b47b8e87c22da42dc1793bec3c38e438ad9bc508e120a16a326

See more details on using hashes here.

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