Element wise multiplication of 2 pandas series containing matrices as elements-CodePudding

Suppose I have 2 pandas series, a and b, where the elements are matrices as follows

a = pd.Series([np.array([[1,2], [3,4]]), np.array([[1,2], [3,4]])])

b = pd.Series([np.array([[1,2], [3,4]]), np.array([[1,2], [3,4]])])

I would like to multiply the series elements of a with b such that the output looks like:

output of series multiplication

I have the following solution which works:


c = []

for i in range(a.shape[0]):

    c.append(np.matmul(a[i], b[i]))

however, this requires a for loop and outputs a list rather than a .Series. Any ideas how this could be done in a more compact form? Thanks.

CodePudding user response：

IIUC use:

c = []

for i in range(a.shape[0]):

    c.append(np.matmul(a[i], b[i]))
    
print (pd.Series(c))

0    [[7, 10], [15, 22]]
1    [[7, 10], [15, 22]]
dtype: object

Or:

print (pd.Series([np.matmul(a[i], b[i]) for i in range(a.shape[0])]))

0    [[7, 10], [15, 22]]
1    [[7, 10], [15, 22]]
dtype: object

CodePudding user response：

Just convert the list to pandas Series and print. Add the following code at the end:

print(pd.Series(c))

CodePudding user response：

I couldn't figure out how to do it in pandas, but it's so much easier if you convert those Series to numpy arrays.

pd.Series(list(np.matmul(np.array(a.to_list()), np.array(b.to_list()))))