Basic plot

Let’s start plotting the empirical data using a histogram and the PDF. These plots will help to visually guide whether a distribution is a good model for a dataset. The confidence intervals are automatically set to 95% CII but can be changed using the alpha parameter during initialization. When using the plot functionality, it automatically shows the histogram in bars and with a line, PDF/CDF, and confidence intervals. All these properties can be manually specified or removed.

We will start generating random data from the normal distribution and create a basic PDF and CDF plot.

# Import
from distfit import distfit
import numpy as np

# Create dataset
X = np.random.normal(0, 2, 10000)
y = [-8,-6,0,1,2,3,4,5,6]

# Initialize
dfit = distfit(alpha=0.01)

# Fit
dfit.fit_transform(X)

# Plot seperately
fig, ax = dfit.plot(chart='pdf')
fig, ax = dfit.plot(chart='cdf')
Basic PDF and CDF plot

figP1a

figP1b

Plot all fitted distributions

# Plot seperately
fig, ax = dfit.plot(chart='pdf', n_top=11)
fig, ax = dfit.plot(chart='cdf', n_top=11)
PDF and CDF plot with multiple fitted distributions.

figP2a

figP2b

Combine plots

# Plot together
fig, ax = dfit.plot(chart='pdf')
fig, ax = dfit.plot(chart='cdf', ax=ax)

# Plot together
fig, ax = dfit.plot(chart='pdf', n_top=11)
fig, ax = dfit.plot(chart='cdf', n_top=11, ax=ax)
Basic PDF and CDF plot

figP3a

figP3b

Change chart properties

# Change or remove properties of the chart.
dfit.plot(chart='pdf',
                pdf_properties={'color': 'r'},
                cii_properties={'color': 'g'},
                emp_properties=None,
                bar_properties=None)

dfit.plot(chart='cdf',
                pdf_properties={'color': 'r'},
                cii_properties={'color': 'g'},
                emp_properties=None)

# Combine the charts and change properties
fig, ax = dfit.plot(chart='pdf',
                pdf_properties={'color': 'r', 'linewidth': 3},
                cii_properties={'color': 'r', 'linewidth': 3},
                bar_properties={'color': '#1e3f5a', 'width': 10})

# Give the previous axes as input.
dfit.plot(chart='cdf',
                n_top=10,
                pdf_properties={'color': 'r'},
                cii_properties=None,
                ax=ax)

# Combine the charts and change properties
fig, ax = dfit.plot(chart='pdf',
                pdf_properties=None,
                cii_properties=None,
                emp_properties={'color': 'g', 'linewidth': 3},
                bar_properties={'color': '#1e3f5a'})

# Give the previous axes as input.
dfit.plot(chart='cdf',
                pdf_properties=None,
                cii_properties=None,
                emp_properties={'color': 'g', 'linewidth': 3},
                ax=ax)
Basic PDF and CDF plot

figP4a

figP4b

figP4c

figP4d

QQ plot

# Plot seperately
fig, ax = dfit.qqplot(X)
fig, ax = dfit.qqplot(X, n_top=11)
Quantile-Quantile plot

figP5a

figP5b

Line plots

# Line plot

# Import
from distfit import distfit
# Initialize
dfit = distfit(smooth=3, bound='up')
# Import
df = dfit.import_example(data='tips')
# Make line plot without any fitting
dfit.lineplot(df['tip'], xlabel='Number', ylabel='Tip value', grid=True, line_properties={'marker':'.'})

# Fit
dfit.fit_transform(df['tip'])
# Create line plot but now with the distribution
dfit.lineplot(df['tip'], xlabel='Number', ylabel='Tip value', grid=True, line_properties={'marker':'.'}, projection=True)
Line plot

figP6a

figP6b