/DataOverview

Library with functionalities similar a "Pandas Dataframe Describe".

Primary LanguageJupyter NotebookMIT LicenseMIT

DataOverview

Library with goal similar to Pandas Dataframe Describe with some additional information.

Current functions:

  • details: This variable return a dataframe with all statistic data. Table on chapter 2, show us a better description of all.
  • columns_type: Return a dictionary that show you a column types distribution.
  • show: This method will filter details variable showing you only columns by type.
    The unique parameter type can be one of follows 'all','int64','float64','object','bool', 'datetime64'.
  • obj_distrib: This method will show the data distribution for column of type 'object'. The parameter 3 possible parameter will better explained on chapter 3.

1 - Usage:

import pandas as pd
import dfview as ovw  

df = pd.read_csv("my_data.csv")
describe = ovw.DataOverview(df)

# getting dictionary of type columns distribuction
cols = describe.columns_type()
print(cols)

# showing describe by type of columns
describe.show(type="int64")
describe.show(type="object")

# showing entire describe 
describe.details
describe.show(type='all')

# getting obj distribution all columns, axis=1, include_nulls=True 
describe.obj_distrib()

# getting obj distribution with some specific columns, axis=0, include_nulls=False 
describe.obj_distrib( columns_list=['col1', 'col2', 'col3'], axis=0, include_nulls=False )

2 - Informations Showed by Variable 'details' or Method 'show'

Column Description
dtype Type of column data.
count Count of non null occurrences.
null Count of null occurrences.
min Smallest value of columns's data. This applies only for int or float column type.
mean Mean of data. This applies only for int or float column type.
max Largest value of column's data. This applies only for int or float column type.
std Standard deviation of data. This applies only for int or float column type.
std% Percentage that represents the size of standard deviation in comparison of data distribution. This applies only for int or float column type.
25% Quantile 25% like a pandas describe method.
50% Quantile 50% like a pandas describe method.
75% Quantile 75% like a pandas describe method.
mode Mode of the column or occurrence data with the most repetitions.
n_mode Occurrence count of the data mode.

3 - Paramteters of obj_distrib

Column Description
column_list List of columns that you want show data distribuction. Default value show all tables.
include_nulls This boolean parameter include or exclude null values on first occurence of dataframe returned. Default value is True.
axis This integer parameter change dataframe disposition. The default value 1 returns columns on top, 0 on left side.