A RetroSearch Logo

Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Search Query:

Showing content from https://www.python-graph-gallery.com/42-custom-linear-regression-fit-seaborn/ below:

Website Navigation


Customize Linear Regression Fit Line Features

Libraries

First, we need to load a few libraries:

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
Dataset

Since scatter plot are made for visualizing relationships between two numerical variables, we need a dataset that contains at least two numerical columns.

Here, we will use the iris dataset that we load directly from the gallery:

path = 'https://raw.githubusercontent.com/holtzy/The-Python-Graph-Gallery/master/static/data/iris.csv'
df = pd.read_csv(path)
df.head()
sepal_length sepal_width petal_length petal_width species 0 5.1 3.5 1.4 0.2 setosa 1 4.9 3.0 1.4 0.2 setosa 2 4.7 3.2 1.3 0.2 setosa 3 4.6 3.1 1.5 0.2 setosa 4 5.0 3.6 1.4 0.2 setosa Color and width

You can custom the appearance of the regression fit in a scatterplot built with seaborn thanks to the line_kws argument.

This argument accepts a dictionary of argument names and their values. These values will be exclusively used to style the line. For example, line_kws={"color" : "red", "linewidth" : 1.5} will render the line red and with a width of 1.5.

fig, ax = plt.subplots(figsize=(8, 6))
sns.regplot(
    x=df["sepal_length"],
    y=df["sepal_width"],
    line_kws={"color": "red", "linewidth": 1.5},
    ax=ax
)
plt.show()
Opacity

You can also custom the opacity of the line with the alpha value:

fig, ax = plt.subplots(figsize=(8, 6))
sns.regplot(
    x=df["sepal_length"],
    y=df["sepal_width"],
    line_kws={
        "color": "r",
        "alpha": 0.4
    },
    ax=ax
)
plt.show()
Line width and style

You can also custom the line width and style with the linewidth and linestyle values:

fig, ax = plt.subplots(figsize=(8, 6))
sns.regplot(
    x=df["sepal_length"],
    y=df["sepal_width"],
    line_kws={
        "color": "darkred",
        "alpha": 0.4,
        "lw": 5,
        "ls": "--"
    },
    ax=ax
)
plt.show()

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4