PandasPlotting Cheat Sheet

A Programming › Python cheat sheet in English by Zayd

Sheet for Pandas Plotting

18,138 views 4,196 downloads 5 Jun 2020 History

PRs

Terms

The terms will be used to refer to:
- df = Pandas DataFrame
- series = Pandas Series
- data = Pandas DataFrame or Series

Plot whith Series and DataFrames

`- series.plot()`	Series
`- dataframe.plot(x='None',y='None')`	DataFrame
`- data.plot.<kind>()`	Another method
`'bar' or 'barh', 'hist', 'box', 'kde' or 'density' , 'hexbin', 'pie' and 'scatter'`	Kinds

Bar Plot

`- data.plot.bar() /.barh()`	Bar plot/ Horizontal plot
`- data.plot.bar(stacked=True)`	Stacked bar plot

Area Plots

`- data.plot.area()`	Area plot
`- data.plot.area(stacked=False)`	Non-Stacked area plot

Pie plot

`- series.plot.pie()`	Pie plot for Series
`- DataFrame.plot.pie(subplots=True)`	Pie plot for DataFrame
`- series.plot.pie(labels= ['A','B','C'], colors= ['r','b','g'], autopct= '%.2f')`	Wedge labels

It's valid:
fontsize and figsize

Scatter plot

`-DataFrame.plot.scatter(x=' ', y=' ')`	Scatter plot
`-ax= df.plot.scatter(x='A',y='B', color='None', label='Group1')`
`-df.plot.scatter(x='C',y='D', color='Other', label='Group2', ax= ax)`	Multiple plot

Histograms

`- data.plot.hist()`	Histogram plot
`- data.plot.hist(stacked=True, bins=10)`	Stacked and bins size
`- data.plot.hist(orientation='horizontal', cumulative=True)`	Horizontal and cumulative
`data.diff().hist(color='g', alpha=0.5)`	Subplots histograms

Box Plots

`- data.plot.box()`	Box plot
`- dict={'boxes':' ','whiskers':' ', 'medians':' ', 'caps': ' '}`	Color of Boxes
`- data.plot.box(color=dict)`
`- data.plot.box(vert=False)`	Horizontal box plot
`- df.boxplot(by='column')`
`- df.boxplot(column=[' ',' '], by=[' ',' ']))`	Groupings
`- df.groupby('g').boxplot()`	^{for random choice}

^{The "choice random" is:}
- g=np.random_choice(['A','B'],size=50])

Hexagonal bin plot

`- DataFrame.plot.hexbin(x='None', y='None')`	Hexagonal bin plot
`- DataFrame.plot.hexbin(x='None', y='None', C= 'z', reduce_C_function=np.max)`	add column 'z' for the value
`- DataFrame.plot.hexbin(x='None', y='None', gridsize= 20)`	Gridsize

Density plot

- data.plot.kde()

Density plot

Plot for data .CSV

> data= pd.read_csv('Name or direction of data')
Andrews curves
- pdt.andrews_curves(data, 'column name with class names')
Parallel coordinates
- pdt.parallel_coordinates(data, 'column name with class names')
RadViz
- pdt.radviz(data, 'column name with class name')

Plotting Tools from Pandas Plotting

> import pandas.plotting as pdt
Scatter matrix plot
- pdt.scatter_matrix('frame', 'alpha= 0.5', 'figsize=(6,6)')
Lag plot
- pdt.lag_plot(series)
Autocorrelation plot
- pdt.autocorrelation_plot(series)
Bootstrap plot
- pdt.bootstrap_plot(series, size= 50, samples= 500, color='green')

Plot formatting

Plot style
- series.plot(style='k--')
Controlling the legend
- DataFrame.plot(legend= False)
Color map
- DataFrame.plot(colormap=' ')
Scales (logarithmic)
- data.plot(logy= True) or logx or loglog
Plotting on a seconday y-axis
- DataFrame.column1.plot()
- DataFrame.column2.plot(secondary_y=True)
Suppressing tick resolution adjustament
- data.plot(x_compat= True)

Subplots

`- data.plot(subplots=True)`	Subplots
`- data.plot(subplots= True, layout= (2,3)`	Multiple axes

It´s valid:
figsize and sharex

Plotting with errors bars

DataFrame.plot.bar(yerr=df_err, xerr= df1_err, capsize=3)

df_err and df1_err are DataFrame of the errors of X and Y

Plotting tables

- ax.get_xaxis().set_visible(False)
- DataFrame.plot(table= True, ax=ax)
Adds table to:
- fig, ax= plt.subplots(1,1)
- pdt.table(ax, DataFrame, loc='upper right', colWidths=[0.2, 0.2, 0.2])
- DataFrame.plot(ax= ax)

j

Badges

Nice Cheat Sheet Downloaded from Unusual Locations × 2

Comments

No comments yet.