How to multiply individual elements of numpy array of row ith with element of another numpy array of row ith?

The inventory example is that I want to multiply an numpy array(containing the item's (280 of them) costing in USD, Euro) of size [280,2] with an numpy array of size [280,3] (stocks in 3 store houses(representing the column).

I believe I have no problem using for loops to calculate but I am trying to learn techniques of broadcasting and reshape. So I would like your help to point me the correct direction(or methods)

Edit: Example

  Array A            
  [[1.50 1.80]        
  [3    8   ]]        

  Array B
  [[5  10 20]
  [10 20 30]]

Result I require is

  [[7.5 9  11.5 18  30 36]
  30  80 60   160 90  240]]

Thanks

Solution

The description was a bit fuzzy, as was the example:

In [264]: A=np.array([[1.5,1.8],[3,8]]); B=np.array([[5,10,20],[10,20,30]])                          
In [265]: A.shape                                                                                    
Out[265]: (2, 2)
In [266]: B.shape                                                                                    
Out[266]: (2, 3)

Looks like you are trying to do a version of outer product, which can be done with broadcasting.

Let's try one combination:

In [267]: A[:,:,None]*B[:,None,:]                                                                    
Out[267]: 
array([[[  7.5,  15. ,  30. ],
        [  9. ,  18. ,  36. ]],

       [[ 30. ,  60. ,  90. ],
        [ 80. , 160. , 240. ]]])

The right numbers are there, but not the right order. Let's try again:

In [268]: A[:,None,:]*B[:,:,None]                                                                    
Out[268]: 
array([[[  7.5,   9. ],
        [ 15. ,  18. ],
        [ 30. ,  36. ]],

       [[ 30. ,  80. ],
        [ 60. , 160. ],
        [ 90. , 240. ]]])

That's better - now just reshape:

In [269]: _.reshape(2,6)                                                                             
Out[269]: 
array([[  7.5,   9. ,  15. ,  18. ,  30. ,  36. ],
       [ 30. ,  80. ,  60. , 160. ,  90. , 240. ]])

_268 is a partial transpose of _267, .transpose(0,2,1).