question regarding of Forest_Degradation_Vogelmann_et_al notebook

[hydrogeohc]

Hi,
I tried to reproduce your workflow for different remote sensing dataset since I cannot import the datacube API successfully. I don't understand this function
def _flatten_shallow(arr):
return arr.reshape(arr.shape[0] * arr.shape[1])

Could it be possible for the author to explain the reshape array function? shape[0] for time dimension?
shape[1] for the NDVI values ?

Thanks!