Let us make confusion matrix for cross validation based results

Let us import a dataframe, first import the needed python libraries

from sklearn.metrics import confusion_matrix
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import cross_val_score
import numpy as np 
import pandas as pd 

Now import the dataframe

df = pd.read_csv('animals_classification.csv')
df

Suppose we have a data frame called df ok when we are working on jupyter notebook or Google colabs. Now split the dataframe into x(the features we feed into the machine learning model), and y(the label or target)

x_my=df.drop(['animal'], axis=1)
x_my

y_my=df['animal']

Then we train the data into a machine learning model for explanation purpose let us assume we used decision trees.

dt = DecisionTreeClassifier()
# I used 10 fold cross validation
scores = cross_val_score(dt, x_my, y_my, cv=10, scoring='accuracy')
print(scores) 
print('Model accuracy score : {0:0.4f}'.format(scores.mean()))

Now let us obtain the matrix of predicted target values and the true target values originally found in the dataset.

y_pred = cross_val_predict(dt, x_my, y_my, cv=10)
conf_mat = confusion_matrix(y_my, y_pred)
print(conf_mat)

Now simply we can feed this into seaborn and get a colored confusion matrix.

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="YlGn", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

Our confusion matrix

Suppose your confusion matrix has 3 classes only, as Dog,Cat, Pig then modify the code as

array=[[a, b , c],
 [ d ,e,  f],
 [ g , h ,i]]

confusion_matrix = pdNew.DataFrame(array, range(3), range(3))
snNew.heatmap(confusion_matrix, annot=True ,cmap="YlGn", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig']);
pltNew.show()

so its easy just modify the range as per the number of target classes you have, and feed in the correct array, and enter the names of target class as needed for the true target names, and predicted target names.

Now for something extra, let us plot confusion matrix in different shades of colours.

flare colored confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="flare", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

summer colored confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="summer", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

Wistia coloured confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="Wistia", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

BuPu coloured confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="BuPu", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

YlOrBr coloured confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="YlOrBr", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

OrRd colored confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="OrRd", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

YlGnBu colored confusion matrix

import seaborn as snNew
import pandas as pdNew
import matplotlib.pyplot as plt     
import matplotlib.pyplot as pltNew

ax= plt.subplot()

array=[[20,  3 , 0, 0 ],
 [ 0 ,5,  13,  0],
 [ 3 , 0 ,2,  18],
 [ 2 , 18 , 3 ,0]]

confusion_matrix = pdNew.DataFrame(array, range(4), range(4))
snNew.heatmap(confusion_matrix, annot=True ,cmap="YlGnBu", ax=ax)

ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.xaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']); ax.yaxis.set_ticklabels(['Dog', 'Cat','Pig','Snake']);
pltNew.show()

Thanks for reading, hope this blog was useful. Have fun coding !!!