CODEWITHSUNDEEP
numpy
Pavan
Pavan

Reputation: 391

appending values to numpy array

I have two different arrays b0 and b1 where: b0=[1,2] b1=[3,4]

I want list[1st element of b0, 1st element of b1] to appended into new array B and similarly: list[2nd element of b0, 2nd element of b1] to appended into new array B and so on......

that is my new array should be something like: array([1,3],[2,4])

Below is my code: 

b0=np.array([1,2])
b1=np.array([3,4])

for val in range(len(b1)):
    L=[b0[val],b1[val]]
    B=np.append(L,axis=0)
print(B)

I am getting missing on positional argument values error. Kindly help me to fix it.

Upvotes: 1

Views: 2386

Answers (3)

hpaulj
hpaulj

Reputation: 231325

In [51]: b0=np.array([1,2]) 
    ...: b1=np.array([3,4])

Order's wrong:

In [56]: np.vstack((b0,b1))                                                  
Out[56]: 
array([[1, 2],
       [3, 4]])

but you can transpose it:

In [57]: np.vstack((b0,b1)).T                                                
Out[57]: 
array([[1, 3],
       [2, 4]])

stack is a more general purpose concatenator

In [58]: np.stack((b0,b1), axis=1)                                           
Out[58]: 
array([[1, 3],
       [2, 4]])

or with:

In [59]: np.column_stack((b0,b1))                                            
Out[59]: 
array([[1, 3],
       [2, 4]])

More details on combining arrays in my other recent answer: https://stackoverflow.com/a/56159553/901925

All these, including np.append use np.concatenate, just tweaking the dimensions in different ways first. np.append is often misused. It isn't a list append clone. None should be used repeatedly in a loop. They make a new array each time, which isn't very efficient.

Upvotes: 0

Xuekai Du
Xuekai Du

Reputation: 717

Using np.append here isn't the most convenient way in my opinion. You can always cast python list into np.array and it's much easier to just use zip in this case. 

b0=np.array([1,2])
b1=np.array([3,4])
B=np.array(list(zip(b0,b1)))

output:

>>> B
array([[1, 3],
       [2, 4]])

Upvotes: 0

N. Arunoprayoch
N. Arunoprayoch

Reputation: 942

If you insist to use numpy array, this is what I would do.

new = []
for x, y in zip(b0, b1):
    new.append([x, y])

new = np.array(new)

Or list comprehension

new = np.array([[x,y] for x, y in zip(b0, b1)])

Result:

array([[1, 3],
       [2, 4]])

Upvotes: 1

Related Questions