## Why numpy converts 20000001 int to float32 as 20000000.?

I am trying to put some numbers into numpy array
>>> np.array([20000001]).astype('float32')
array([ 20000000.], dtype=float32)
where did 1 go?
...

more
»