Menu
Perhaps this has been raised and addressed somewhere else but I haven't found it. Suppose we have a numpy array:
a = np.arange(100).reshape(10,10)
b = np.zeros(a.shape)
start = np.array([1,4,7]) # can be arbitrary but valid values
end = np.array([3,6,9]) # can be arbitrary but valid values
start and end both have valid values so that each slicing is also valid for a.
I wanted to copy value of subarrays in a to corresponding spots in in b:
b[:, start:end] = a[:, start:end] #error
this syntax doesn't work, but it's equivalent to:
b[:, start[0]:end[0]] = a[:, start[0]:end[0]]
b[:, start[1]:end[1]] = a[:, start[1]:end[1]]
b[:, start[2]:end[2]] = a[:, start[2]:end[2]]
I wonder if there is a better way of doing this instead of an explicit for-loop over the start and end arrays.
Thanks!
628
Slice Two-dimensional Numpy Arrays To slice elements from two-dimensional arrays, you need to specify both a row index and a column index as [row_index, column_index] . For example, you can use the index [1,2] to query the element at the second row, third column in precip_2002_2013 .
One-Dimensional SlicingThe first item of the array can be sliced by specifying a slice that starts at index 0 and ends at index 1 (one item before the 'to' index). Running the example returns a subarray with the first element. We can also use negative indexes in slices.
To select an element from Numpy Array , we can use [] operator i.e. It will return the element at given index only.
Numpy with PythonContents of ndarray object can be accessed and modified by indexing or slicing, just like Python's in-built container objects. As mentioned earlier, items in ndarray object follows zero-based index. Three types of indexing methods are available − field access, basic slicing and advanced indexing.
We can use broadcasting to create a mask of places to be edited with two sets of comparisons against start and end arrays and then simply assign with boolean-indexing for a vectorized solution -
# Range array for the length of columns
r = np.arange(b.shape[1])
# Broadcasting magic to give us the mask of places
mask = (start[:,None] <= r) & (end[:,None] >= r)
# Boolean-index to select and assign
b[:len(mask)][mask] = a[:len(mask)][mask]
Sample run -
In [222]: a = np.arange(50).reshape(5,10)
...: b = np.zeros(a.shape,dtype=int)
...: start = np.array([1,4,7])
...: end = np.array([5,6,9]) # different from sample for variety
...:
# Mask of places to be edited
In [223]: mask = (start[:,None] <= r) & (end[:,None] >= r)
In [225]: print mask
[[False True True True True True False False False False]
[False False False False True True True False False False]
[False False False False False False False True True True]]
In [226]: b[:len(mask)][mask] = a[:len(mask)][mask]
In [227]: a
Out[227]:
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
[20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
[30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49]])
In [228]: b
Out[228]:
array([[ 0, 1, 2, 3, 4, 5, 0, 0, 0, 0],
[ 0, 0, 0, 0, 14, 15, 16, 0, 0, 0],
[ 0, 0, 0, 0, 0, 0, 0, 27, 28, 29],
[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With
