Creating a scatter plot by category in python

Creating scatter plots for has become so easy with python. For datasets with a manageable number of categories, manual mapping with a dictionary is recommended. It is intuitive, gives explicit control, and is easy to extend or debug

The dataset used in this code snippet can be found in vincentarelbundock.github.io/Rdatasets/datasets.html by scrolling down to datasets category(datasets/iris.csv)

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.patches import Patch

iris= pd.read_csv('iris.csv')

#assiging each category of species a color
Dict_colors = {'setosa':'blue','versicolor':'green','virginica':'red'}

#map species to their respective colors
colors = iris['Species'].map(Dict_colors)

#creating a scatter plot
plt.scatter(iris['Petal.Length'],iris['Sepal.Length'],c=colors)

#showing the grid
plt.grid(True, linestyle = '--', alpha = 0.5)

#creating a custom legend using matplotlib.patches

legend_patches = [Patch(color=color, label= species) for species, color in Dict_colors.items()]
plt.legend(handles=legend_patches, title='species',loc='upper left')

#setting labels grid and title
plt.gca().set(xlabel= 'petal length',ylabel='sepal lenght', title= 'Scatter plot of sepal length vs petal length by species')

Code Explanation

1. Importing Required Libraries python

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.patches import Patch

pandas: Used for reading and manipulating the dataset (iris.csv).

matplotlib.pyplot: Provides functions for creating and customizing visualizations.

matplotlib.patches.Patch: Allows the creation of custom legend elements, specifically colored boxes for categorical data (e.g., species).

2. Loading the Dataset

iris = pd.read_csv('iris.csv')

Purpose: Reads the Iris dataset (iris.csv) into a DataFrame named iris.

pd.read_csv(): Loads a CSV file into a DataFrame.

The dataset typically contains columns such as:

Sepal.Length: Length of the sepal (in cm).

Petal.Length: Length of the petal (in cm).

Species: Categorical column with three classes (setosa, versicolor, virginica).

3. Assigning Colors to Species

Dict_colors = {'setosa': 'blue', 'versicolor': 'green', 'virginica': 'red'}

Creates a dictionary mapping each species to a specific color.

'setosa': Blue.

'versicolor': Green.

'virginica': Red.

This ensures that each species is represented consistently in the scatter plot.

4. Mapping Colors to Data Points

colors = iris['Species'].map(Dict_colors)

Purpose: Maps the Species column in the dataset to its corresponding color from Dict_colors.

map(): A pandas method that applies a function or mapping (dictionary in this case) to a Series.

If the Species value is 'setosa', it assigns 'blue'.

Similarly, 'versicolor' gets 'green', and 'virginica' gets 'red'.

• Result: A Series of color values (blue, green, red) matching the species.

5. Creating the Scatter Plot

plt.scatter(iris['Petal.Length'], iris['Sepal.Length'], c=colors)

plt.scatter(): Plots points on a 2D plane.

x-axis: iris['Petal.Length'] (petal length values).

y-axis: iris['Sepal.Length'] (sepal length values).

c=colors: Colors the points based on the mapped species colors.

Visualization: Each point represents a flower, and its position is determined by petal and sepal length, while its color indicates the species.

6. Showing the Grid

plt.grid(True, linestyle='--', alpha=0.5)

Purpose: Adds a grid to the plot to make it easier to read.

Parameters:

True: Enables the grid.

linestyle='--': Dashed grid lines.

alpha=0.5: Adjusts transparency of the grid lines for a cleaner look.

7. Creating a Custom Legend

legend_patches = [Patch(color=color, label=species) for species, color in Dict_colors.items()]
plt.legend(handles=legend_patches, title='species', loc='upper left')

Purpose: Adds a custom legend to explain the color-coding of species.

Patch: Creates colored boxes that match the colors in the scatter plot.

color: The color of the box (e.g., blue, green, red).

label: The species name (e.g., 'setosa').

handles: The list of Patch objects is passed to the legend.

title: Adds a title (species) to the legend.

loc='upper left': Positions the legend in the upper-left corner of the plot.

8. Setting Labels and Title

plt.gca().set(xlabel='petal length', ylabel='sepal length', title='Scatter plot of sepal length vs petal length by species')

plt.gca(): Retrieves the current axes object (the plot’s coordinate system).

.set(): Sets multiple properties of the axes:

xlabel: Label for the x-axis ('petal length').

ylabel: Label for the y-axis ('sepal length').

title: Title of the plot ('Scatter plot of sepal length vs petal length by species').