I have a numpy matrix with 130 X 13. Say I want to select a specific set of rows meeting a condition and a subset of columns - trainx[trainy==label,[0,6]] The above code does not work and throws an error - IndexError: shape mismatch: indexing arrays could not be broadcast together with shapes (43,) (2,). However if I do it in 2 steps - first subset rows and then columns, it works. Is it something weird or numpy works this way? temp1 = trainx[trainy==label,:] temp1 = temp1[:,[0,6]] Continue reading...
