Online Matplotlib Compiler

Matplotlib draws charts from Python code: line and bar charts, scatter plots, histograms, pie charts and many more. It saves them as PNG, SVG or PDF files, and pandas and seaborn draw their charts with it. People use it for figures in reports, papers and homework, and to explore data. This page is an online Matplotlib compiler: the code runs in your browser, so you can try it without installing anything. Run the example first, then paste any snippet below into a new cell to try it.

What the example does

np.random.seed(0) fixes the random numbers, so every run draws the same chart. np.random.randn(1000) gives 1,000 values from a standard normal distribution for each axis. np.random.randint(10, 101, size=1000) gives 1,000 whole numbers from 10 to 100, once for the colours and once for the sizes. plt.scatter() draws one marker per point: c=colors with cmap="viridis" maps each number to a colour from dark purple to yellow, s=sizes sets each marker's area in points squared, and alpha=0.5 makes the markers half transparent, so overlapping points stay visible. plt.xlabel(), plt.ylabel() and plt.title() add the text, and plt.show() displays the chart below the cell.

Draw a line chart with a legend

Most Matplotlib code starts with plt.subplots(), which returns a figure and an axes to draw on. Give each line a label, and ax.legend() lists them:

import matplotlib.pyplot as plt

months = ["Jan", "Feb", "Mar", "Apr", "May", "Jun"]
online = [120, 135, 150, 145, 170, 190]
in_store = [200, 180, 175, 160, 150, 140]

fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(months, online, marker="o", label="Online")
ax.plot(months, in_store, marker="s", linestyle="--", color="gray", label="In store")
ax.set_title("Monthly sales")
ax.set_xlabel("Month")
ax.set_ylabel("Units sold")
ax.legend()
ax.grid(alpha=0.3)
plt.show()

figsize is the width and height in inches. marker, linestyle and color change how each line looks.

Put several charts in one figure

plt.subplots(1, 3) returns one row of three axes, and each can hold a different chart type:

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(seed=1)
heights = rng.normal(170, 8, size=500)

fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(12, 4))

bars = ax1.bar(["A", "B", "C", "D"], [23, 17, 35, 29], color="tab:green")
ax1.bar_label(bars)
ax1.set_ylim(0, 40)
ax1.set_title("Bar chart")

ax2.hist(heights, bins=20, edgecolor="white")
ax2.set_title("Histogram")

ax3.pie([45, 30, 25], labels=["Rent", "Food", "Other"], autopct="%1.0f%%")
ax3.set_title("Pie chart")

fig.suptitle("Three chart types")
plt.show()

bar_label() writes each bar's value on top of it, and autopct prints each slice's share. plt.subplots(2, 2) returns a grid instead, and axes[0, 1] is its top-right chart.

Change the look with a style

A style sets colours, backgrounds and grid lines. plt.style.available lists the built-in ones, and plt.style.context() applies one only inside the with block:

import matplotlib.pyplot as plt

print(plt.style.available)

with plt.style.context("ggplot"):
    fig, ax = plt.subplots(figsize=(6, 4))
    ax.plot([1, 2, 3, 4, 5, 6], [2, 4, 3, 6, 5, 8], marker="o", label="Team A")
    ax.plot([1, 2, 3, 4, 5, 6], [1, 3, 4, 4, 6, 7], marker="o", label="Team B")
    ax.set_title("ggplot style")
    ax.legend()
    plt.show()

plt.style.use("ggplot") applies it to every chart after it, until you call plt.style.use("default").

Save a figure as PNG, SVG or PDF

fig.savefig() writes the figure to a file, and the extension sets the format. dpi sets the resolution of a PNG, and bbox_inches="tight" trims the empty margin. The files appear in the sidebar's Data tab:

import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(6, 4))
ax.bar(["Q1", "Q2", "Q3", "Q4"], [120, 135, 150, 170])
ax.set_title("Quarterly sales")

fig.savefig("sales.png", dpi=200, bbox_inches="tight")
fig.savefig("sales.svg", bbox_inches="tight")
fig.savefig("sales.pdf")
plt.show()

Good to know

• On this page, plt.savefig() clears the figure once the file is written, so a plt.show() after it displays an empty chart. fig.savefig() leaves the figure as it is.
• Pyplot keeps drawing on the same figure until it is shown. A cell that calls plt.plot() without plt.show() shows [<matplotlib.lines.Line2D object at ...>] instead of a chart, and that line then turns up in the next chart you show.
• Only the fonts that ship with Matplotlib, such as DejaVu Sans, are available here. They have no Chinese, Japanese or Korean characters, so such text is drawn as empty boxes, with a warning that the glyphs are missing.