#Data Visualization

import seaborn as sns

import matplotlib.pyplot as plt

sns$.$barplot$($x$=$"class", y$=$data$["$class$"].$index, palette$=$'mako', data$=$mushroom$\_$data$)$

#The number of poisonous mushrooms is almost twice the number of normal mushrooms. There is an imbalance data problem.

#We will be using Matplotlib pyplot and Seaborn to plot our data.

#$\%\%$

from sklearn import preprocessing

#Label encoding is used to convert categorical features to numerical values.

def label$\_$encode$\_$fit$($mushroom$\_$data, columns$)$:

result $=$ mushroom$\_$data.copy$()$

encoders $=\{\}$

for column in columns:

encoder $=$ preprocessing.LabelEncoder$()$

result$[$column$]=$ encoder.fit$\_$transform$($result$[$column$])$

encoders$[$column$]=$ encoder

return result, encoders

#$\%\%$

data$1,$ encoders$1=$ label$\_$encode$\_$fit$($data$,$data.columns$)$

data$1.$head$(10)$

#$\%\%$

def correlation$\_$map$($mushroom$\_$data, method$)$:

corr $=$ mushroom$\_$data.corr$($method$)$

ix $=$ corr.sort$\_$values$(\text{'}$class$\text{'},$ ascending$=$False$).$index

df$\_$sorted$\_$by$\_$correlation $=$ mushroom$\_$data.loc$[$:$,$ ix$]$

corr $=$ df$\_$sorted$\_$by$\_$correlation.corr$($method$)$

plt$.$subplots$($figsize$=(18,14))$

with sns$.$axes$\_$style$("$white$")$:

# display a correlation heatmap

ax $=$ sns$.$heatmap$($corr$,$ annot$=$True$)$

plt$.$show$()$

#$\%\%$

correlation$\_$map$($data$1,$ method$=$"spearman"$)$

#Gill$\_$size has the highest correlation with class. It should be included to the model.

#There some highly correlated variables such as$,$ gill$-$color & ring$-$type, gill$-$color & bruises, bruises & stalk$-$surface$-$below$-$ring etc. These highly correlated variables ohuld be discarded from the model to obtain more accurate results.

#$\%\%$

y $=$ data$1[[\text{'}$class$\text{'}]]$ # contains only "class", target, variable.

X $=$ data$1.$iloc$[$:$,1$:$]$ # contains independent variable.

#$\%\%$

from sklearn.feature$\_$selection import SelectKBest

import numpy as np

def SelectKBestCustomized$($mushroom$\_$data, k$,$ score$\_$func, target$=$"class"$)$:

X$=$mushroom$\_$data.drop$($columns$=$target$)$

y$=$mushroom$\_$data$[$target$]$

np$.$random.seed$(123)$ # for mutual$\_$info regression

fs $=$ SelectKBest$($score$\_$func$=$score$\_$func, k$=$k$)$

fs$.$fit$($X$,$ y$)$

mask $=$ fs$.$get$\_$support$()$

selected$\_$features $=[$feature for bool, feature in zip$($mask$,$ X$.$columns$)$ if bool$]$

return selected$\_$features

#$\%\%$

from sklearn.feature$\_$selection import mutual$\_$info$\_$classif

mutual$\_$info$\_$classif$($X$,$ y$,$ random$\_$state$=123)$

#$\%\%$

mutual$\_$info$\_$selection $=$ SelectKBestCustomized$($data$1,9,$ mutual$\_$info$\_$classif$)$

#$\%\%$

mutual$\_$info$\_$selection

#$\%\%$

X$\_$new $=$ X$[[\text{'}$odor$\text{'},$'gill$-$size',

'gill$-$color',

'stalk$-$surface$-$above$-$ring',

'stalk$-$surface$-$below$-$ring',

'stalk$-$color$-$above$-$ring',

'stalk$-$color$-$below$-$ring',

'ring$-$type',

'spore$-$print$-$color'$]]$

#$\%\%$

data$\_$selected$\_$features $=$ data$1[[\text{'}$odor$\text{'},$

'gill$-$size',

'gill$-$color',

'stalk$-$surface$-$above$-$ring',

'stalk$-$surface$-$below$-$ring',

'stalk$-$color$-$above$-$ring',

'stalk$-$color$-$below$-$ring',

'ring$-$type',

'spore$-$print$-$color',

'class'$]]$

#$\%\%$

a $=5$ # number of rows

b $=3$ # number of columns

c $=1$ # initialize plot counter

fig $=$ plt$.$figure$($figsize$=(14,22))$

for i in data$\_$selected$\_$features:

plt$.$subplot$($a$,$ b$,$ c$)$

#plt$.$title$(\text{'}\{\},$ subplot: $\{\}\{\}\{\}\text{'}.$format$($i$,$ a$,$ b$,$ c$))$

plt$.$xlabel$($i$)$

sns$.$barplot$($x$=$i$,$ y$=$data$\_$selected$\_$features$[$i$].$index, palette$=$'Set$3\_$r$\text{'},$ hue$=$"class", data$=$data$\_$selected$\_$features$)$

c $=$ c $+1$

plt$.$show$()$ THE PYTHON CODE GIVEN ABOVE IS RELATED TO RANDOM FOREST CLASSIFICATION IN THE DATA SCIENCE COURSE.

PLEASE INTERPRET THIS CODE AND PREPARE A REPORT ACCORDING TO THE SUBJECTS AND CODES.