Flatten numpy array but also keep index of value positions?

前端 未结 9 1262
北荒
北荒 2021-02-04 13:58

I have several 2D numpy arrays (matrix) and for each one I would like to convert it to vector containing the values of the array and a vector containing each row/column index.

相关标签:
9条回答
  • 2021-02-04 14:11

    You can try this using itertools

    import itertools
    import numpy as np
    import pandas as pd
    
    def convert2dataframe(array):
        a, b = array.shape
        x, y = zip(*list(itertools.product(range(a), range(b))))
        df = pd.DataFrame(data={'V':array.ravel(), 'x':x, 'y':y})
        return df
    

    This works for arrays of any shape, not necessarily square matrices.

    0 讨论(0)
  • 2021-02-04 14:11

    Like @miguel-capllonch I would suggest using np.ndindex which allows you to create the desired output like this:

    np.array([(v, *i) for (i, v) in zip(np.ndindex(x.shape), x.ravel())])
    

    which results in an array that looks like this:

    array([[ 3.  0.  0.]
           [ 1.  0.  1.]
           [ 4.  0.  2.]
           [ 1.  1.  0.]
           [ 5.  1.  1.]
           [ 9.  1.  2.]
           [ 2.  2.  0.]
           [ 6.  2.  1.]
           [ 5.  2.  2.]])
    

    Alternatively, using only numpy commands

    np.hstack((list(np.ndindex(x.shape)), x.reshape((-1, 1))))
    
    0 讨论(0)
  • 2021-02-04 14:16

    You could also let pandas do the work for you since you'll be using it in a dataframe:

    x = np.array([[3, 1, 4],
                  [1, 5, 9],
                  [2, 6, 5]])
    df=pd.DataFrame(x)
    #unstack the y columns so that they become an index then reset the
    #index so that indexes become columns.
    df=df.unstack().reset_index()
    df
    
       level_0  level_1  0
    0        0        0  3
    1        0        1  1
    2        0        2  2
    3        1        0  1
    4        1        1  5
    5        1        2  6
    6        2        0  4
    7        2        1  9
    8        2        2  5
    
    #name the columns and switch the column order
    df.columns=['x','y','V']
    cols = df.columns.tolist()
    cols = cols[-1:] + cols[:-1]
    df = df[cols]
    df
    
       V  x  y
    0  3  0  0
    1  1  0  1
    2  2  0  2
    3  1  1  0
    4  5  1  1
    5  6  1  2
    6  4  2  0
    7  9  2  1
    8  5  2  2
    
    0 讨论(0)
  • 2021-02-04 14:25

    The class np.ndindex is especially meant for this, and easily does the trick. Similar efficiency to the np.mesgrid method above, but it requires less code:

    indices = np.array(list(np.ndindex(x.shape)))
    

    For the dataframe, do:

    df = pd.DataFrame({'V': x.flatten(), 'x': indices[:, 0], 'y': indices[:, 1]})
    

    If you don't need the dataframe, just do list(np.ndindex(x.shape)).

    Note: don't get confused between x (the array at hand), and 'x' (the name of the second column).

    I know this question was posted a very long time ago, but just in case it's useful to anyone, as I didn't see np.ndindex being mentioned.

    0 讨论(0)
  • 2021-02-04 14:25

    I am resurrecting this because I think I know a different answer that is way easier to understand. Here is how I do it:

    xn = np.zeros((np.size(x), np.ndim(x)+1), dtype=np.float32)
    row = 0
    for ind, data in np.ndenumerate(x):
        xn[row, 0] = data
        xn[row, 1:] = np.asarray(ind)
        row += 1
    

    In xn we have

    [[ 3.  0.  0.]
     [ 1.  0.  1.]
     [ 4.  0.  2.]
     [ 1.  1.  0.]
     [ 5.  1.  1.]
     [ 9.  1.  2.]
     [ 2.  2.  0.]
     [ 6.  2.  1.]
     [ 5.  2.  2.]]
    
    0 讨论(0)
  • 2021-02-04 14:30

    Another way:

    arr = np.array([[3, 1, 4],
                    [1, 5, 9],
                    [2, 6, 5]])
    
    # build out rows array
    x = np.arange(arr.shape[0]).reshape(arr.shape[0],1).repeat(arr.shape[1],axis=1)
    # build out columns array
    y = np.arange(arr.shape[1]).reshape(1,arr.shape[0]).repeat(arr.shape[0],axis=0)
    
    # combine into table
    table = np.vstack((arr.reshape(arr.size),x.reshape(arr.size),y.reshape(arr.size))).T
    print(table)
    
    0 讨论(0)
提交回复
热议问题