PLEASE HELP IN PYTHON

Bagging and boosting are considered ensemble methods of machine learning. Ensemble is a concept in which multiple models, also called decision trees, are trained using the same learning algorithm to produce a better predictive performance than using a single decision tree. The ensembles take part in a bigger group of methods, called multi$-$classifiers, where a set of hundreds or thousands of learners with a common objective are fused together to solve the problem.

For this given code, you implement either a bagging or a boosting version to solve the problem of a pharmaceutical drug analysis. Fill out the missing code sections titled "Write Your Code Here". One section has an error stating: "DataFrame.drop$()$ takes from $1$ to $2$ positional arguments but $3$ were given". There is an excel file provided for this code. Also, try to explain each section of the code on how it works if you can. Thank you.

import pandas as pd

import seaborn as sns #for plotting

from sklearn.ensemble import RandomForestClassifier #for the model

from sklearn.tree import DecisionTreeClassifier

from sklearn.tree import export$\_$graphviz #plot tree

from sklearn.model$\_$selection import train$\_$test$\_$split #for data splitting

np$.$random.seed$(123)$ #ensure reproducibility

#Let's load the data. This is clinical data to predict wether a patient is likely to suufer a heart attack $(1)$ or not $(0).$ For each patient $(303$ in total$)14$ variables have been taken into account for the prediction. We will use all these data to train a randon forest as a predictor for heart attacks:

df $=$ pd$.$read$\_$csv$(\text{'}$heart$2.$csv$\text{'})$

print$($df$.$shape$)$

df$.$head$()$

#Let's have a better look at the data $($visualization$)$:

X$=$np$.$array$($df$[$df$.$columns$[0$:$14]])$

f$,$a $=$ plt$.$subplots$(4,4)$

f$.$set$\_$figheight$(10)$

f$.$set$\_$figwidth$(10)$

a $=$ a$.$ravel$()$

for idx,ax in enumerate$($a$)$:

if idx$<mn>1</mn><mn>4</mn>$:

ax$.$hist$($list$($X$[$:$,$idx$]),$color$=\text{'}$r$\text{'})$

ax$.$set$\_$title$($df$.$columns$[$idx$])$

plt$.$tight$\_$layout$()$

#Now, let's train our random forest. Not all available data will be used for training, some will be used for testing $(20\%)$:

X$\_$train, X$\_$test, y$\_$train, y$\_$test $=$ train$\_$test$\_$split$($df$.$drop$(\text{'}$target$\text{'},1),$ df$[\text{'}$target$\text{'}],$ test$\_$size $=\mathrm{.}2,$ random$\_$state$=10)$

model $=$ RandomForestClassifier$($max$\_$depth$=5)$

model.fit$($X$\_$train, y$\_$train$)$

#There is a type error on this sections stating: DataFrame.drop$()$ takes from $1$ to $2$ positional arguments but $3$ were given

#Now, let's predict the class $($likely to have an attak $(0)$ or not $(1))$ for our testing data of patients, using our trained random forest:

y$\_$predict $=$ model.predict$($X$\_$test$)$

#These are the predicted labels by the random forest:

y$\_$predict

#$...$and these are the actual labels of the testing data:

y$\_$test$=$np$.$array$($list$($y$\_$test$))$

y$\_$test

#Let's see the difference between the predicted and the actual labels. See that, out of $61$ predictions, the random forest mistook $12$:

sum$($abs$($y$\_$predict $-$ y$\_$test $))$

#Now, for boosting, let's train five random forests:

#Write your code here

let's see the predictions for each random forest:

#Write your code here

#Now, our final prediction will be an average of the predictions of the five random forests. If more than two random forests say that the label must be $1,$ then, accordingly, the final prediction will be $1$ and $0$ otherwise:

#Write your code here

#See that our final $($averaged$)$ predictions have only $3$ mistakes, this is$,$ the boosting does improve indeed the predictions of our initial single model $(12$ mistakes$)$:

sum$($abs$($y$\_$predict $-$ np$.$array$($final$)))$