Answer the following $5$ NumPy related tasks for data analysis. These will require use of NumPy functions and

methods, matrix manipulations, vectorized computations, NumPy statistics, NumPy where function, etc. To

complete this task, write a function called task$1($data$,$ filter$\_$value, type$\_$of$\_$card$),$ where

CITS $2401$

Computer Analysis

and Visualisation

Page $2$ of $6$

data contains all records from the dataset and filter$\_$value is an area name and type$\_$of$\_$card is the

name of the card provider. The function should return a list containing values from the following questions.

Return all results rounded to two decimal points.

Input:

cos$\_$dist, var, median, corr, pca $=$ task$1($data$,$ 'Port Lincoln $\text{'}$

$,$

'ANZ $\text{'})$

output:

$[0\mathrm{.}06,1337142\mathrm{.}45,[5\mathrm{.}75,7\mathrm{.}21],-0\mathrm{.}06,[0\mathrm{.}73,0\mathrm{.}81,0\mathrm{.}7,0\mathrm{.}93,0\mathrm{.}72,...]]$

i$.$ cos$\_$dist: Calculate cosine distance between normal and malicious transactions based on

IP$\_$validity$\_$score.

Formula:

$=1$

$(,)=$

Output:

print$($cos$\_$dist$)=0\mathrm{.}06$

ii$.$ var: Filter transactions based on certain geographical area e$.$g$.$ Port Lincoln and calculate and display

the variance of transaction amount for a specific area. Note: use the Actual area column from the

dataset. Use sample variance formula for calculation.

Output:

print$($var$)=1337142\mathrm{.}45$

iii. median: Filter data based on Type

$\_$

of

$\_$

card and then calculate the median of

Authentication$\_$score value for transactions that are in the lower $25$th $($inclusive$)$ and upper

$75$th $($inclusive$)$ percentile.

Output:

print$($median$)=[5\mathrm{.}75,7\mathrm{.}21]$