In Python:

$1.$ Load a $.$mat file containing a data set

$2.$ Answer how many linear discriminant functions this data set can have

$3.$ Partition the data set into $80\%$ training set and $20\%$ testing set

$4.$ Fit the training partition using both LDA and PCA

$5.$ Then project both training and testing in $1,2,3,...,19$ dimensions $($x$-$axis$)$ to $5$ nearest neighbor accuracy $($y$-$axis$).$

$6.$ Plot using scatter plot the training and testing for both LDA and PCA in both $2$d scatterplot and $3$d scatter plot.

$7.$ Comment on the difference between the $2$d and $3$d plots along with LDA vs PCA.

$8.$ Ensure that the code compiles, produces the correct plots and is properly named with x$-$label, y$-$label, and z$-$label. Also have a legend for the different classes in the dataset