In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
is there a function in verde to extract the points from grid based on lists of coordinates?

In reply toundefined

In reply toundefined

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
There is a sample method. Something like this. coords = [(x, y) for x, y in zip(survey.X, survey.Y)] survey['sample_points'] = [x[0] for x in grid.sample(coords)] . There is an example in the docs somewhere but I couldn't find it.

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
Many thanks Craig Miller. Will test that. Thsi sample method should work also with general xarray grid? Or sepcific only to Verde?

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
in that case it was a geosoft or surfer grid I read in using rasterio

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
but I think an xarray grid will work to. I think it's actually a rasterio method come to think of it.

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
rather than verde

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
https://rasterio.readthedocs.io/en/latest/api/rasterio.sample.html

In reply to@_slack_fatiando_UMGLPTLAW:matrix.org
but I was using it in a script with a bunch of verde stuff as well.

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
ok thanks. I tried with a standard `xarray.Dataset`but it seems that the the sample methods do not exist in xarray

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
I found a solutions but it works only with lang, lot coordinates

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
in any case what i need is a little more complicated since i need to finde the nearest coordinate of my survey in a grid, so I think the sample method will not works…

In reply to@_slack_fatiando_UMSRSPEMA:matrix.org
That's an xarray thing and it should have a method to interpolate values on a set of points. Can't find it now but look into "indexing"

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
ok thnask

In reply to@_slack_fatiando_UMSRSPEMA:matrix.org
Found it: http://xarray.pydata.org/en/stable/indexing.html#nearest-neighbor-lookups

python
data = vd.grid_to_table(grid)
interp = vd.ScipyGridder(method="nearest")
interp.fit((data.easting, data.northing), data.field)
sample_points = (easting, northing)
sample_field = interp.predict(sample_points)

In reply to@_slack_fatiando_UMSRSPEMA:matrix.org
You can also use Verde for this:

python
data = vd.grid_to_table(grid)
interp = vd.ScipyGridder(method="nearest")
interp.fit((data.easting, data.northing), data.field)
sample_points = (easting, northing)
sample_field = interp.predict(sample_points)

In reply to@_slack_fatiando_UMSRSPEMA:matrix.org
And then you can choose which interpolation method you want to do the sampling. ScipyGridder supports nearest neighbours, linear, and cubic. You could also use a vd.Spline there or any other gridder.

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
leouieda yes sure I forgot the predict option…

In reply to@_slack_fatiando_UT0BDHKPE:matrix.org
good points…will test this approach