Example - Interpolate Missing Data
[1]:
import rioxarray # for the extension to load
import xarray
%matplotlib inline
Load in xarray dataset
[2]:
xds = xarray.open_dataarray("MODIS_ARRAY.nc")
[3]:
xds
[3]:
<xarray.DataArray (y: 200, x: 200)>
array([[ nan, nan, nan, ..., 656., 656., 554.],
[ nan, nan, nan, ..., 694., 694., 642.],
[ nan, nan, nan, ..., 456., 575., 642.],
...,
[993., 817., 817., ..., 471., 479., 498.],
[893., 893., 816., ..., 479., 479., 469.],
[816., 816., 832., ..., 515., 469., 485.]], dtype=float32)
Coordinates:
* y (y) float64 5.05e+06 5.05e+06 5.05e+06 ... 5.004e+06 5.004e+06
* x (x) float64 -7.274e+06 -7.274e+06 ... -7.228e+06 -7.228e+06
Attributes:
crs: +a=6371007.181 +b=6371007.181 +lon_0=0 +no_defs +proj=sinu +u...
res: [231.65635826 231.65635826]
is_tiled: 0
nodata: -28672.0
transform: [ 2.31656358e+02 0.00000000e+00 -7.27400965e+06 0.00000000e...
[4]:
xds.isel(x=slice(0, 20), y=slice(0, 20)).plot()
[4]:
<matplotlib.collections.QuadMesh at 0x7f3bf197f978>
Fill missing with interpolate_na
API Reference:
DataArray: rio.interpolate_na()
Dataset: rio.interpolate_na()
[5]:
filled = xds.rio.interpolate_na()
[6]:
filled
[6]:
<xarray.DataArray (y: 200, x: 200)>
array([[673., 558., 687., ..., 656., 656., 554.],
[673., 558., 558., ..., 694., 694., 642.],
[673., 558., 558., ..., 456., 575., 642.],
...,
[993., 817., 817., ..., 471., 479., 498.],
[893., 893., 816., ..., 479., 479., 469.],
[816., 816., 832., ..., 515., 469., 485.]], dtype=float32)
Coordinates:
* y (y) float64 5.05e+06 5.05e+06 5.05e+06 ... 5.004e+06 5.004e+06
* x (x) float64 -7.274e+06 -7.274e+06 ... -7.228e+06 -7.228e+06
spatial_ref int64 0
Attributes:
transform: (231.6563582639561, 0.0, -7274009.649486291, 0.0, -231.656...
_FillValue: -28672.0
grid_mapping: spatial_ref
[7]:
filled.isel(x=slice(0, 20), y=slice(0, 20)).plot()
[7]:
<matplotlib.collections.QuadMesh at 0x7f3bf1865860>
