Show Menu
Cheatography
  1. Home >
  2. Cheat Sheets >

Pandas Less Frequent Functions Cheat Sheet (DRAFT) by

This is a draft cheat sheet. It is a work in progress and is not finished yet.

Read and Write

import pandas as pd
 
pd.rea­d_c­sv(­fil­epa­th_­or_­buffer, sep=', ', names=­None, index_­col­=None)
 
df.to_­csv­(pa­th_­or_buf, sep, column­s=None, header­=True, index=­True, index_­lab­el=­None, mode='w')

Numerical Features and Nans

df.sort_values(
by=feat, 
ascending=False)
sort table according to the values of a columns
 
df[feat].dropna()
remove lines with a NaN value
 
df[feat].isna()
check if the value of feat is NaN in the table

Catego­rical Features

pd.get­_du­mmi­es(df[feature])
Transform a
categorical feature
into dummy variables

Features Visual­ization

df[feats].plot­(ki­nd=­['d­ensity' | 'bar'], subplo­ts=­True, layout=(1, 2), sharex=, figsize=);
distib­ution of numeric features
 

General Infos and Basic Statistics

df.info()
general infos
 
df.des­cribe()
basic statistics
on numerical features
 
df.describe(include=[
'object', 'bool'])
include non-nu­merical features

Apply Functions

df.apply(my_function)
# ex: df.app­ly(­lambda x: )
apply a function
 
df['feat'] = df['feat'].map(d)
# or
df = df.replace({'feat': d})
replace values in a column according to dict d

Group by

df.gro­upby(['feat'])
[columns_to_keep].func()
group by a feature
 
df.gro­upby([feat])
[columns_to_keep].
agg([list_of_functions])
group by a feature and apply several functions

Cross Tables

pd.cro­sst­ab(df['feat1'], 
df['feat2'], normal­ize=)
confusion matrix
 
df.pivot_table(
['featur­es_­to_­analyze'],
['groupi­ng_feat'],
aggfunc='mean')
pivot table