Please solve this ERROR!!!!!!!

KeyError Traceback $($most recent call last$)$

in $()$

$90$

$91$ # Build the decision tree

$--->92$ decision$\_$tree $=$ build$\_$decision$\_$tree$($df$)$

$93$

$94$ # Print the decision tree

$1$ frames

in build$\_$decision$\_$tree$($data$,$ tree$)$

$78$ # Set the outcome for the current branch

$79$ if len$($counts$)==1$:

$--->80$ tree$[$best$\_$split$\_$attribute$][$value$]=$ outcomes$[0]$

$81$ else:

$82$ tree$[$best$\_$split$\_$attribute$][$value$]=$ outcomes$[$index$]$

KeyError: 'House Type'

I checked my csv file but House Type data is well contained in the csv file.

$============================$

import pandas as pd

import numpy as np

# Load dataset into a pandas dataframe

df $=$ pd$.$read$\_$csv$(\text{'}$dataset$.$csv$\text{'})$

# Define a function to calculate entropy

def entropy$($target$\_$col$)$:

elements, counts $=$ np$.$unique$($target$\_$col, return$\_$counts$=$True$)$

probs $=$ counts$/$len$($target$\_$col$)$

entropy $=$ np$.$sum$(-$probs $*$ np$.$log$2($probs$))$

return entropy

# Define a function to calculate information gain

def info$\_$gain$($data$,$ split$\_$attribute$\_$name, target$\_$name$=$"Outcome"$)$:

# Calculate the entropy of the entire dataset

total$\_$entropy $=$ entropy$($data$[$target$\_$name$])$

# Calculate the values and corresponding counts for the split attribute

vals, counts $=$ np$.$unique$($data$[$split$\_$attribute$\_$name$],$ return$\_$counts$=$True$)$

# Calculate the weighted entropy of the split data

weighted$\_$entropy $=$ np$.$sum$([($counts$[$i$]/$np$.$sum$($counts$))*$ entropy$($data$.$where$($data$[$split$\_$attribute$\_$name$]==$vals$[$i$]).$dropna$()[$target$\_$name$])$ for i in range$($len$($vals$))])$

# Calculate the information gain

info$\_$gain $=$ total$\_$entropy $-$ weighted$\_$entropy

return info$\_$gain

# Define a function to get the best split attribute

def get$\_$best$\_$split$($data$)$:

# Get the list of column names

columns $=$ list$($data$.$columns$)$

# Remove the target column name

columns.remove$(\text{'}$Outcome$\text{'})$

# Calculate the information gain for each column

info$\_$gains $=[$info$\_$gain$($data$,$ column$)$ for column in columns$]$

# Get the index of the column with the highest information gain

best$\_$column$\_$index $=$ np$.$argmax$($info$\_$gains$)$

# Return the name of the best split attribute

return columns$[$best$\_$column$\_$index$]$

# Define the decision tree building function

def build$\_$decision$\_$tree$($data$,$ tree$=$None$)$:

# Get the best split attribute

best$\_$split$\_$attribute $=$ get$\_$best$\_$split$($data$)$

# Get the unique values for the best split attribute

values $=$ np$.$unique$($data$[$best$\_$split$\_$attribute$])$

# Create a new tree node with the best split attribute

if tree is None:

tree $=\{\}$

tree$[$best$\_$split$\_$attribute$]=\{\}$

# For each value of the best split attribute, create a new branch

for value in values:

# Create a new branch for the current value

sub$\_$data $=$ data.where$($data$[$best$\_$split$\_$attribute$]==$ value$).$dropna$()$

# Get the most common outcome for the current branch

outcomes, counts $=$ np$.$unique$($sub$\_$data$[\text{'}$Outcome$\text{'}],$ return$\_$counts$=$True$)$

index $=$ np$.$argmax$($counts$)$

# Set the outcome for the current branch

if len$($counts$)==1$:

tree$[$best$\_$split$\_$attribute$][$value$]=$ outcomes$[0]$

else:

tree$[$best$\_$split$\_$attribute$][$value$]=$ outcomes$[$index$]$

# Recursively build the subtree for the current branch

if len$($sub$\_$data.drop$($columns$=[$best$\_$split$\_$attribute, 'Outcome'$]))>0$:

subtree $=$ build$\_$decision$\_$tree$($sub$\_$data.drop$($columns$=[$best$\_$split$\_$attribute$]),\{\})$

tree$[$best$\_$split$\_$attribute$][$value$]=$ subtree

# Return the tree

return tree

# Build the decision tree

decision$\_$tree $=$ build$\_$decision$\_$tree$($df$)$

# Print the decision tree

decision$\_$tree