来自numpy数组的Python等高线图

Question

I have a 2D numpy array made from zeros and ones that I use as a mask for other arrays. I was trying to use matplotlib.contour to highlight an area on a plot, but every time I try it I get a zero-size array to minimum.reduce without identity error. Any idea?

Since this mask is a set of rectangles, I tried to find the edges manually, but it does not work properly. Here's the code I use:

tmp1,tmp2 = [],[]
for ii in range(len(mask))[1:-2]:
    if mask[ii+1] - mask[ii] != 1: tmp1.append(mask[ii])
if mask[ii] - mask[ii-1] != 1: tmp2.append(mask[ii]-1)


rect_limits = []
for ii in range(len(tmp1)):
    rect_limits.append([- delta_cont, tmp1[ii], delta_cont, tmp2[ii]])

that way tmp1 and tmp2 should give me the max and min of the rectangles I am searching for. (the lateral edges of the rectangle are fixed, so no problem there).

then I just need to use add_patch to create the contour of the rectangles I want.

Any alternative idea to find the rectangle edges?

Edit:

OK, so my mask would be something like:

[[0 0 0 0 0 0 0 0 0 0 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0],      
[0 0 0 0 0 0 1 1 1 1 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0],
[0 0 0 0 0 0 0 0 0 0 0 0],
[0 0 0 0 0 0 0 0 0 0 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0],
[0 0 0 0 0 0 1 1 1 1 0 0]]

and ideally what I would like as a result would be:

[[1,3],[6,9]]

ie, an array built with

[[y_start1,y_end1],[y_start2,y_end2],...]

Answer 1

answer.py

mygrid = [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
          [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0],
          [0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0]]

def findEdges(grid)
    y_start = -1
    saved = []
    for lineno, row in enumerate(grid):
        # Case where we don't have a start point
        if y_start == -1 and 1 in row:
            y_start = lineno
        # Case where we have a start point and we just hit a zero row
        if y_start != -1 and 1 not in row:           
            saved.append((y_start, lineno-1))
            y_start = -1
        # Case where we have a start point and hit the end of the table
        if lineno == len(grid)-1 and y_start > 0:
            saved.append((y_start, lineno))

    return saved

print(findEdges(mygrid))

This gives an output of:

mike@example ~ $ python answer.py
[(1, 3), (6, 9)]

Note: This won't work if two or more rectangles are allowed to be side by side on the grid.