Change Figure Size when using df.plot() in Pandas

df.plot() is a quick way to chart a DataFrame, but the default figure size is small and often cropped. The good news is that Pandas plotting wraps Matplotlib, so every Matplotlib figure-size trick works on it.

Quick Example

Pass figsize=(width, height) straight to df.plot(). Sizes are in inches.

import pandas as pd

df.plot(figsize=(10, 6))

The figsize argument appears on every Pandas plotting method, from df.plot.line() to df.plot.bar(), ss.plot.hist(), and the rest.

The figsize argument

The most direct approach is the figsize keyword on df.plot(). A (10, 6) figure gives the axes enough room for readable labels and a legend without spilling over.

ax = df.plot(figsize=(10, 6))

Compare the small default with the sized-up version:

Pandas df.plot with the small default figure size

Pandas df.plot with figsize set to 10 by 6 inches

Pandas passes figsize through to Matplotlib, so the value is always a (width, height) tuple in inches.

Size a plot you create yourself

If you build the figure first, set the size there and hand the axes to df.plot(). This is the pattern to use when a notebook already has a Matplotlib figure you want to reuse.

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(10, 6))
df.plot(ax=ax)

Passing ax=ax tells Pandas to draw into your pre-sized axes instead of creating its own. The figsize then lives in one place: the plt.subplots() call.

Resize subplots

For a stacked set of charts, Pandas splits the axes with subplots=True, and figsize controls the whole grid while layout controls its shape.

df.plot(
    subplots=True,
    figsize=(9, 8),
    layout=(2, 2),
    sharex=True,
)

Pandas df.plot with subplots=True using a figsize and 2 by 2 layout

The single figsize sizes the entire canvas of panels. Count the number of columns in your DataFrame and pick a layout that fits them reasonably.

Set a global default with rcParams

To make every figure in a script the same size without repeating figsize, set Matplotlib's default once at the top.

import matplotlib.pyplot as plt

plt.rcParams["figure.figsize"] = (10, 6)

df.plot()

Pandas df.plot sized by a global rcParams figure.figsize default

Setting plt.rcParams["figure.figsize"] changes the default for all subsequent figures, including every df.plot() that does not pass its own figsize. Any explicit figsize argument still overrides it.

Practical Tips

Use df.plot(figsize=(w, h)) for a quick, one-off resize.
Build the figure with plt.subplots(figsize=...) and pass ax=ax when you want full control or reuse.
Set plt.rcParams["figure.figsize"] once when every chart should share the same size.
Combine subplots=True with figsize and layout to size a grid of panels.
Use bbox_inches="tight" when saving so rotated or long labels are not clipped.

Python Charts

Python plotting and visualization demystified