ExtractByMask - Script to include pixels whose center falls outside of mask

BenjaminBigarre

Hi, 

I would like to know if there is a way to extract all pixels of a raster touched by a mask, not just the pixels whose center falls within the mask, even if it requires some scripting. 

I have been trying to clip a raster with a mask (tried raster and feature), however some pixels are excluded from the output since their center fall outside of the mask (see picture). I cannot change the pixel size of the input raster. 

I know gdalwarp has an option to keep all pixels that touch the mask, is there a way to code something similar with arcpy?

I attached my code for reference. 

Benjamin

Moorea_ARCPY.py

Screenshot 2024-04-16 132304.png

SusanJarvis_1-1733251333985.png

SusanJarvis_0-1733251290059.png

BenjaminBigarre

Here is my solution to create a customized mask from raster, that covers the entire surface of interest: 

1. Convert raster to points
2. Identify points of cells overlapping my surface of interest
3. Select points, export to a table
4. Use a custom function to extract cells of raster with list of points, and create the raster mask. 

Below is my code, screenshot of final result on arcgis is attached

import arcpy
from arcpy import env as e,SpatialReference
import os
from arcpy import da as d
from arcpy.sa import *
import pandas as pd
import numpy as np
from arcgis.features import *

projGDB = r"C:\Users\bigar\Documents\ArcGIS\Projects\MooreaHydro2024\MooreaHydro2024.gdb"
e.workspace = r"C:\Users\bigar\Documents\ArcGIS\Projects\MooreaHydro2024\MooreaHydro2024.gdb"
e.overwriteOutput = True
pd.set_option('display.max_rows', None)

def createmyMask(inRaster,pointlist):
    ## Creates a raster mask by extracting raster cells by coordinates.
    desc=arcpy.Describe(inRaster)
    rastermask=ExtractByPoints(in_raster=inRaster, points=pointlist,extraction_area="INSIDE")
    rastermask.save("C:/Users/bigar/Documents/ArcGIS/Projects/MooreaHydro2024/MooreaHydro2024.gdb/"+desc.baseName+"Mask")
    desc2=arcpy.Describe(rastermask)
    print("Mask {} was created".format(desc2.baseName))

def extractXY(gdbTable):
    df = pd.DataFrame(GeoAccessor.from_table(gdbTable))
    # df = df.astype({'X': int, 'Y':int})
    coordlist=[]
    for index, row in df.loc[:, ['X', 'Y']].iterrows():
        # print(type(int(row['X'])), type(int(row['Y'])))
        coordlist.append(arcpy.Point(int(row['X']),int(row['Y'])))
        print(coordlist)
    return coordlist

if __name__ == "__main__":
    XYlist = extractXY(r"C:\Users\bigar\Documents\ArcGIS\Projects\MooreaHydro2024\MooreaHydro2024.gdb\Opunohu_MODIS_CELL_COORD")

    createmyMask("C:/Users/bigar/Documents/ArcGIS/Projects/MooreaHydro2024/MooreaHydro2024.gdb/MOD16A3GF_061_ET_500m_doy2004001_aid0001_reproj", XYlist)

 

Screenshot 2024-04-18 151045.png