如何在 Python 中动态连接二维数组？

Question

I am trying to concatenate 2D arrays horizontally using numpy in a dynamic way. Starting from an empty array I want to add 2D arrays depending on the if condition outcome. I don't know the final dimension so I cannot define anything before the loop starts. Assuming I have the 2D arrays called A,B,C:

X = np.array([])
for name in modules:
   if name = 'AAA':
      X = np.append(X,A, axis = 1)
   if name = 'BBB'
      X = np.append(X,B, axis = 1)
   if name = 'CCC'
      X = np.append(X,C, axis = 1)

After reading how np.append works I realized why this solution is wrong. Is there an easy way to produce concatenation on the fly? N:B: I know the number of columns and rows of A,B,C (they have the same number of rows) but I cannot know how many matrices will be concatenated since all is depending on the if conditions. The concatenation order is important and should be as reported in the code.

Answer 1

You could create a dictionary which functions as a lookup-table, matching the names with the arrays.

mapper = {'AAA': A,
          'BBB': B,
          'CCC': C}

X = np.hstack([mapper[name] for name in modules])

Answer 2

Because at first, X does not have the same number of rows as A , B or C , you can not append X with any of them. Here is a small tweak:

X = None
for name in modules:
   if name = 'AAA':
      X = np.append(X,A, axis = 1) if X is not None else A
   if name = 'BBB'
      X = np.append(X,B, axis = 1) if X is not None else B
   if name = 'CCC'
      X = np.append(X,C, axis = 1) if X is not None else C

Hope this would be helpful.