Assignment $1$: K$-$nearest neighbors classifier $($KNN$)$

Mandatory. You are required to submit Python code with extension $".$py$"$ that includes an implementation of a KNN classifier and a test on it$.$ You can use the attached template knn$.$py $($including a minimum test$)$ and fill the unimplemented parts there.

The submitted code must include comments that explain what the code does for every functional block $($you don't have to write comments line$-$by$-$line$).$

Hint. It may be useful to know that a vector X $($of numpy.ndarray$)$ can be sorted according to values in a vector d $($of numpy. ndarray$)$ by: X $=$ X$[$d$.$argsort$()]$

where d$.$argsort$()$ returns a vector n such that n$[$i $]$ is an index at which the value d$[$i$]$ should be in sorting d; see the documentation of numpy. argsort for detail. You may use this tip for sorting K training data points X according to the distances d from them to a given input.

This is the template code that I want you to fill out and the goal of it is to get an accuracy of over $70\%.$

class KNearestNeighborsClassifier:

$$ # K: the number of training data points joining the voting

$$ def $\_\_$init$\_\_($self$,$ K$)$:

$$ self.K $=$ K

$$ # fit: train this model on training inputs X and outputs Y

$$ # X: training inputs $--$ np$.$ndarray

$$ # $($shape: $[$# of data points, # of features$])$

$$ # Y: training outputs $--$ np$.$ndarray

$$ # $($shape: $[$# of data points$])$

$$ def fit$($self$,$ X$,$ Y$)$:

$$ # TODO: IMPLEMENT ME

$$ pass

$$ # predict: classify given data points

$$ # X: inputs to the classifier $--$ np$.$ndarray

$$ # $($shape: $[$# of data points, # of features$])$

$$ def predict$($self$,$ X$)$:

$$ # TODO: IMPLEMENT ME

$$ # Hint: Euclid distances between the training inputs X$\_$ and

$$ # $$ prediction inputs X with shape $[$# of data points, # of features$]$ are

$$ # $$ calculuated by $``$numpy$.$sqrt$((($X$\_-$ X$)**2.).$sum$($axis$=1))``$

$$ pass

# check this is a main file

if $\_\_$name$\_\_==\text{'}\_\_$main$\_\_\text{'}$:

$$ import numpy as np

$$ from sklearn.datasets import load$\_$iris

$$ from sklearn.model$\_$selection import train$\_$test$\_$split

$$ iris$\_$dataset $=$ load$\_$iris$()$

$$ X$\_$train, X$\_$test, Y$\_$train, Y$\_$test $=$ train$\_$test$\_$split$($iris$\_$dataset.data,

$$ iris$\_$dataset.target,

$$ random$\_$state$=0)$

$$ knn $=$ KNearestNeighborsClassifier$(3)$

$$ knn$.$fit$($X$\_$train, Y$\_$train$)$

$$ X$\_$test$\_$predict $=$ knn$.$predict$($X$\_$test$)$

$$ accuracy $=$ np$.$sum$($Y$\_$test $==$ X$\_$test$\_$predict$)/$ Y$\_$test.shape$[0]$

$$ print$(\text{'}$acc:$\text{'},$ accuracy$)$