Added catchment delineation functions and exposed to Python #53
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@joeperdefloep Nice work! This would be a great addition to the package. |
|
ooh,quick reply! some other important notes I forgot to mention previously:
EDIT: so this is caused by the line extending to the edge of the grid... If I make a grid of 6 and then draw a line until y1=4, there is no problem, but an only 0 line. I think this may need some more work...
|
|
reference for better line-drawing, including slopes >1: https://www.geeksforgeeks.org/bresenhams-line-generation-algorithm/ |
feefladder
force-pushed
the
delineate
branch
2 times, most recently
from
c45532c
to
e96a122
Compare
feefladder
force-pushed
the
delineate
branch
4 times, most recently
from
8289e25
to
c51ebaf
Compare
|
Ok... So I found out that my commit was 45+ commits behind the main branch, which made some things obsolete that I wrote before, and gave some headaches while fixing. It all works now! import richdem as rd
import numpy as np
dem = rd.rdarray(np.array([
[1,2,3,4,5],
[1,2,3,4,5],
[1,2,3,4,5],
[1,2,3,4,5],
[1,2,3,4,5],]),no_data=0)
mask = rd.rdarray(np.array([
[0,0,0,0,0],
[0,0,0,0,0],
[1,0,0,0,0],
[0,0,0,0,0],
[0,0,0,0,0]]),no_data=0)
print(rd.UpslopeCells(dem,mask=mask,method="Quinn"))
print(rd.UpslopeCells(dem,mask=mask,method="mflow"))output: but also a lot of these when I start richdem: and a segfault at the end.
full output:
I think this has something to do with my installation, rather than the PR
|
|
@giswqs Since it didn't get replies or merged quickly, I also implemented it in Cython, albeit less generic: %%cython -+
from libcpp.queue cimport queue
from libcpp.vector cimport vector
from libc.math cimport isnan
cdef struct xypoint:
long x
long y
# 3 2 1
# 4 X 0
# 5 6 7
cdef int[8] dx = [1, 1, 0, -1, -1, -1, 0, 1]
cdef int[8] dy = [0, 1, 1, 1, 0, -1, -1, -1]
def upslope_cells(double[:,:] mask, double[:,:] dem) -> void:
cdef queue[xypoint] qu
for i in range(mask.shape[0]):
for j in range(mask.shape[1]):
if mask[i,j] != 0:
qu.push(xypoint(i,j))
while not qu.empty():
point = qu.front()
qu.pop()
for i in range(8):
x = point.x + dx[i]
y = point.y + dy[i]
if x < 0 or x > mask.shape[0]:
continue
if y < 0 or y > mask.shape[1]:
continue
if mask[x,y] == 0:
if dem[x,y] > dem[point.x,point.y]:
qu.push(xypoint(x,y))
mask[x,y] += 1hope it helps! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Added catchment delineation, because I needed it.
edit: Actually, this PR finds all upslope cells, given an input, using multiple possible methods. This is different than catchment delineation, because two catchments can actually overlap if a divergent flow metric is used. That functionality is actually already present in a modified algorithm for depression-filling
I saw that it was a todo, so took the liberty to implement it.
I needed the functionality myself to also use a mask in stead of a line.
some notes, in no particular order:
catchment_delineation.hppfunctions. This is actually important, because the raster should be filled with nodata values in order for the algorithm to work properly. I also do this there, but not super nice yet, because I wrote a 2-liner helper function that only works for 3/4 cases0cells, to give more user flexibility,flowProportionsmethod was fast enough for what I needed. Can add on request. (Also I do not see an easy way to implement Dinf without copying a lot of code)FlowProportions, which includes a redundantcopyFromWrapped, but this seemed easier/more understandable than re-writing.It could be that some documentation for c++ is not fully optimal yet, but I can work on that. For now I just wanted to make this pull request. EDIT: documentation should be rather good.
hope you like it!