[FIXED] How do I get a mean INLCUDING NaN values in Python?

Issue

Apparently I have the opposite of everyone else's problem... I would like to take the mean of a pandas dataframe, and I would like to have the result return NaN if there are ANY NaNs in the frame. However, it seems like neither np.mean nor np.nanmean do this. Example code:

b = pd.DataFrame([[1,2],[math.nan,4]])
print(b)
print(np.mean(b))
print(np.nanmean(b))

Result:

Expected Result:

Solution

It can be done using pandas DataFrame.mean() function :

b.mean(axis=None,skipna=False)

According to the documentation of numpy, the reason that you don't get nan from mean function is that when numpy sees the input of mean() is not multiarray.ndarray, it tries to use mean function from that data type if possible.

if type(a) is not mu.ndarray:
    try:
        mean = a.mean
    except AttributeError:
        pass
    else:
        return mean(axis=axis, dtype=dtype, out=out, **kwargs)

In your case it is like calling the mean() function of pandas with axis=None. Which is like running the following code:

b.mean(axis=None) # 2.333333

If you are insisting on using numpy.mean() to calculate mean, you can convert your DataFrame to numpy array before callin np.mean().

array = b.to_numpy()
np.mean(array) # nan

Answered By - Mohsen_Fatemi