cannot replace with np.nan

[ankurk017]

a=np.random.randint(15,size=[2,5])
b=np.random.random([2,5])
b[0,2]=np.nan
a[0,2]=np.nan
I can easily replace any indices of b with nan. But if the same command I am trying to repeat for a, it's showing error. Why? How do I resolve that? Why is this happen?

[gontajones]

Numpy treat NaN as a float. You cannot use it in an integer array.
Instead, you can use an invalid value:

a[0, 2] = -99999

[volcano63]

Insert this line after assignments - and it will become clear

print(a.dtype, type(np.nan))

Though exception - which you should have posted - explains it rather clearly

Error:
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-10-363eed6533b3> in <module>()
----> 1 a[0,2] = np.nan

ValueError: cannot convert float NaN to integer

[carlhyde]

ValueError: cannot convert float NaN to integer

The ValueError: cannot convert float NaN to integer raised because of Pandas doesn't have the ability to store NaN values for integers. From Pandas v0.24, introduces Nullable Integer Data Types which allows integers to coexist with NaNs. This does allow integer NaNs . This is the pandas integer, instead of the numpy integer.

df['column_name'].astype(np.float).astype("Int32")