Python Scatter Plot Tutorial: matplotlib, seaborn, and plotly (2026)

Scatter plots are the workhorse of scientific figures: two continuous variables, one relationship. But making a scatter plot that is honest, readable, and journal-ready takes more than a single ax.scatter() call. This guide walks through the most useful variations in matplotlib, with notes on when each one applies.

1. Basic scatter plot

Start with ax.scatter(x, y). Set s (point size) and use white edgecolors to separate overlapping points slightly.

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
x = rng.normal(5, 1.5, 80)
y = 1.2 * x + rng.normal(0, 1.0, 80)

fig, ax = plt.subplots(figsize=(6, 5))
ax.scatter(x, y, color="#6b21a8", s=40, edgecolors="white", linewidths=0.4)
ax.set_xlabel("Variable X")
ax.set_ylabel("Variable Y")
ax.set_title("Basic Scatter Plot")
plt.tight_layout()

2. Encode a third variable with a colour map

Pass a numeric array to c and a cmap name. Always add a colourbar with a label so readers can decode the mapping.

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
x = rng.uniform(0, 10, 120)
y = rng.uniform(0, 10, 120)
z = np.sin(x) * np.cos(y)   # third dimension encoded as color

fig, ax = plt.subplots(figsize=(6, 5))
sc = ax.scatter(x, y, c=z, cmap="plasma", s=50, edgecolors="none")
plt.colorbar(sc, ax=ax, label="sin(x)·cos(y)")
ax.set_xlabel("X")
ax.set_ylabel("Y")
plt.tight_layout()

3. Add a regression line

Overlay a least-squares fit with np.polyfit. Report slope, intercept, and R² either in the legend or in the figure caption.

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
x = rng.normal(5, 1.5, 80)
y = 1.2 * x + rng.normal(0, 1.0, 80)

m, b = np.polyfit(x, y, 1)
xline = np.linspace(x.min(), x.max(), 200)

fig, ax = plt.subplots(figsize=(6, 5))
ax.scatter(x, y, color="#6b21a8", s=40, edgecolors="white", linewidths=0.4, label="Data")
ax.plot(xline, m * xline + b, color="#f59e0b", lw=2, label=f"y = {m:.2f}x + {b:.2f}")
ax.set_xlabel("Variable X")
ax.set_ylabel("Variable Y")
ax.legend(frameon=False)
plt.tight_layout()

4. Error bars on scatter points

For experimental data with measurement uncertainty, use ax.errorbarinstead of ax.scatter. Set fmt to control the marker style and capsize for the bar cap length.

import matplotlib.pyplot as plt
import numpy as np

x   = np.array([1, 2, 3, 4, 5])
y   = np.array([2.1, 3.8, 6.2, 8.1, 10.3])
xerr = np.array([0.2, 0.3, 0.2, 0.4, 0.3])
yerr = np.array([0.5, 0.4, 0.6, 0.5, 0.7])

fig, ax = plt.subplots(figsize=(6, 5))
ax.errorbar(x, y, xerr=xerr, yerr=yerr,
            fmt="o", color="#6b21a8", ecolor="gray",
            elinewidth=1.5, capsize=4, ms=7)
ax.set_xlabel("Variable X")
ax.set_ylabel("Variable Y")
plt.tight_layout()

Common mistakes

• Overplotting — when many points land on top of each other, use alpha=0.3 or switch to a hexbin plot.
• Unlabelled axes — always include the variable name and its units on both axes.
• Trendline without uncertainty — show a 95% confidence band around regression lines in small samples.
• Misleading aspect ratio — stretching one axis exaggerates or hides correlation.