Regression line plot

Seaborn basics

2 min read

Published Oct 7 2025

ChartsGraphsMatplotlibNumPyPandasPythonSeabornVisualisation

seaborn.regplot() plots data points (scatter) along with a regression line (best-fit line).
It is an axes-level function (draws on a single subplot).

It’s mainly used to visualise the relationship between two numeric variables, showing both the trend (linear regression) and the data spread.

Syntax:

sns.regplot(

data=None,

x=None,

y=None,

x_estimator=None,

x_bins=None,

x_ci='ci',

scatter=True,

fit_reg=True,

order=1,

robust=False,

logx=False,

ci=95,

n_boot=1000,

seed=None,

line_kws=None,

scatter_kws=None,

color=None,

marker='o',

ax=None,

**kwargs

)

Parameters:

data = DataFrame containing the data
x, y = Numeric variables for regression
scatter = If True, shows scatter points
fit_reg = If True, draws regression line
order = Degree of polynomial (1 = linear)
robust = Use robust regression (less sensitive to outliers)
logx = If True, log-transform the x-axis
ci = Confidence interval around regression line (in %)
line_kws = Dictionary of keyword args for the line (e.g., colour, linestyle)
scatter_kws = Dictionary of keyword args for scatter points
color = Base colour for both scatter and line
ax = Axis to plot on (for subplot use)

Basic example

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip")

plt.show()

Plots a scatterplot with a best-fit linear regression line. The shaded area shows the 95% confidence interval.

Turn off scatter or line

Line Only:

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", scatter=False)

plt.show()

Shows only the regression line.

Scatter Only:

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", fit_reg=False)

plt.show()

Shows only the scatter points (same as sns.scatterplot()).

Change line appearance

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(

data=tips,

x="total_bill",

y="tip",

line_kws={"color": "red", "linestyle": "--", "linewidth": 2}

)

plt.show()

Customises line colour, style, and width.

Add scatter customisation

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(

data=tips,

x="total_bill",

y="tip",

scatter_kws={"s": 80, "alpha": 0.6, "color": "green"}

)

plt.show()

Adjusts point size (s), transparency (alpha), and colour.

seaborn reg plot scatter customise example

Polynomial regression (nonlinear)

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", order=2, color="purple")

plt.show()

Fits a 2nd-degree polynomial (curve instead of straight line). You can use higher orders (e.g., order=3) for more curvature.

Robust regression

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", robust=True)

plt.show()

Uses a robust estimator to reduce the effect of outliers on the regression line.

Logarithmic x-axis

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", logx=True)

plt.show()

Applies a log transformation to the x-axis — useful for skewed data.

Change confidence interval

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", ci=68)

plt.show()

Shows a 68% confidence interval instead of the default 95%. Set ci=None to remove the shaded band completely.

Binned regression (x_bins)

import seaborn as sns

import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

sns.regplot(data=tips, x="total_bill", y="tip", x_bins=10)

plt.show()

Groups data along the x-axis into bins, showing average y-values for each bin. Useful for very large datasets or when data are not evenly distributed.