05.a sns.clustermap basic sample

This script shows how to use seaborn.clustermap to plot a heatmap with a colored bar annotating the columns. While clustermap is designed for hierarchical clustering, this example deactivates that feature to maintain the original data order. Its main purpose is to demonstrate how to visually group columns by passing a color mapping to the col_colors parameter.

Note

The hierarchical clustering has been deactivated.

• 
• 

Out:

     A0  A1  A2  A3  A4  A5  A6  A7  A8  A9  A10  A11  A12  A13  A14  A15  A16  A17  A18  A19
A0    1   1   0   8   1   8   0   9   3   2    3    2    9    3    5    7    2    5    5    6
A1    7   4   6   9   4   9   2   7   8   9    7    6    3    0    3    5    1    8    6    3
A2    8   1   3   5   2   9   6   4   2   9    2    4    4    0    9    4    0    1    7    6
A3    5   7   4   1   1   2   9   3   8   5    1    3    0    3    3    1    7    2    2    3
A4    1   8   4   7   9   6   5   9   4   2    4    9    3    4    3    5    5    6    0    0
A5    0   2   1   6   7   7   0   9   4   6    2    0    0    7    4    0    4    1    9    2
A6    2   5   5   6   6   8   1   8   6   9    0    2    3    5    7    8    9    9    6    4
A7    9   3   9   8   0   2   7   4   9   3    7    0    9    3    7    2    5    8    4    8
A8    3   9   9   1   8   6   7   0   2   9    8    7    2    0    2    3    9    9    7    2
A9    5   1   3   5   4   1   9   2   8   5    9    5    3    3    8    0    8    3    9    7
A10   8   0   7   9   9   2   2   6   9   9    1    5    3    2    9    2    4    7    5    4
A11   0   8   8   6   6   3   8   5   5   2    1    4    8    6    5    5    7    9    6    7
A12   7   6   0   7   9   1   3   9   7   0    3    7    1    7    9    8    2    4    8    6
A13   0   7   0   8   2   4   4   2   3   0    7    9    2    7    6    7    5    8    3    3
A14   5   1   6   6   9   1   9   3   5   8    4    4    5    5    5    4    3    5    7    4
A15   8   2   1   3   7   0   0   8   8   9    6    0    5    9    2    5    6    9    8    4
A16   9   3   0   6   9   3   6   3   7   4    1    3    9    5    1    9    5    4    3    1
A17   5   0   5   0   5   3   1   7   7   4    9    7    8    3    1    6    0    8    2    0
A18   9   2   8   1   9   8   5   3   2   9    5    9    6    6    1    3    4    0    9    9
A19   2   6   9   9   5   6   8   2   2   8    1    8    7    8    2    6    4    0    6    9
['b', 'g', 'b', 'r', 'g', 'g', 'b', 'r', 'r', 'b', 'b', 'b', 'g', 'r', 'g', 'g', 'g', 'g', 'r', 'r']
A0     Group b
A1     Group g
A2     Group b
A3     Group r
A4     Group g
A5     Group g
A6     Group b
A7     Group r
A8     Group r
A9     Group b
A10    Group b
A11    Group b
A12    Group g
A13    Group r
A14    Group g
A15    Group g
A16    Group g
A17    Group g
A18    Group r
A19    Group r
dtype: object
C:\Users\kelda\Desktop\repositories\github\python-spare-code\main\examples\matplotlib\plot_main05_a_clustermap.py:88: UserWarning:

FigureCanvasAgg is non-interactive, and thus cannot be shown

"""
# Display 0
iris = sns.load_dataset("iris")
species = iris.pop("species")
lut = dict(zip(species.unique(), "rbg"))
row_colors = species.map(lut)
"""

# Libraries
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

# Constants
N = 20
C = sns.color_palette("Spectral", n_colors=5, as_cmap=False)

# Create data
data = np.random.randint(low=0, high=10, size=(N, N))
colors = [np.random.choice(['r', 'g', 'b']) for i in range(N)]
series = pd.Series({'A%s'%i:'Group %s'%c for i,c in enumerate(colors)})

# Create DataFrame
df = pd.DataFrame(data,
    index=['A%s'%i for i in range(N)],
    columns=['A%s'%i for i in range(N)]
)

# Show
print(df)
print(colors)
print(series)

# Display 1
# ---------
# Create colors dictionary
col_colors = dict(zip(df.columns, colors))

# Show
g = sns.clustermap(df,
    figsize=(5,5),
    row_cluster=False, col_cluster=False,
    #row_colors=col_colors,
    col_colors=pd.Series(col_colors),
    linewidths=0,
    xticklabels=False, yticklabels=False,
    center=0, cmap="vlag"
)

# Display 2
# ---------
# Create colors dictionary

col_colors = series.map({
    'Group g': 'g',
    'Group b': 'b',
    'Group r': 'r'
})

# Show
g = sns.clustermap(df,
    figsize=(5,5),
    row_cluster=False, col_cluster=False,
    #row_colors=col_colors,
    col_colors=pd.Series(col_colors),
    linewidths=0,
    xticklabels=False, yticklabels=False,
    center=0, cmap="vlag"
)

# Show
plt.show()