Assignment plot confusion

i want to know what 2 4 5 means here

They give you the code that was used to generate the graphs. You can see that they define the following array at the beginning:

indices = [2, 4, 5]

You can see that they then use indices to sample the corresponding elements from various data items. So you could pick different index values and see if the graphs look much different or not.

1 Like

it is correct [2,4,5] are indeed images,
for example, 2 is the index
they used 2 to get features of the image from the real dataset with index 2
and also to get the features of the image from the fake dataset with index 2
so the meaningful images in the pairplot are only the diagonal images where it compares the real and fake distribution
the other plots in pair plot don’t give any knowledge since they compare different images
(this is my understanding is it correct )

But I think the point is that there is no relation between the real and fake images. That is to say, there is no reason to believe that fake[2] is any more related to real[2] than it is to real[5] or real[42], right? So all the entries have equal information at least in my understanding. Now the question is why the “off diagonal” elements render differently in the plot as shown.