X is a 100x2 vector that contains sepal and pethal length data for 2 kinds of flowers. y is a 100x1 vector that contains label values for my data: -1 and 1. I create a meshgrid, then I plotted my meshgrid using countourf method and now I load the data in the meshgrid with this code:
for idx, cl in enumerate (np.unique(y)):
plt.scatter (x=X[y == cl, 0], y= X[y == cl, 1], alpha=0.8, c=colors[idx], marker= markers [idx], label = cl, edgecolor = 'black')
alpha, colors, marker, edgecolor are just secondary things. Also np.unique(y) = [-1 1] .
My question is, Why [y==cl,0] is a false and [y==,1] is a true argument? and, How using == can I classifier my data?
In the for loop, cl loops over the two unique values of y i.e. [-1, 1]. So in the first loop iteration, cl=-1 and y == cl returns the rows where your y values are -1. Therefore, [y == cl, 0] returns the locations where y is -1 and the data is the first column (index 0) which is Sepal length. Similarly, [y == cl, 1] returns the locations where y is -1 and the data is the second column (index 1) which is Petal length.
The same applies in the second iteration of the loop when cl is 1.