Yes, but with a lot of information, dask is the only option, and working well with the index. https://github.com/dask/dask/issues/958

El jue., 9 abr. 2020 a las 2:54, Xin Zhang (notifications@github.com) escribió:

@JavierRuano https://github.com/JavierRuano Nice suggestion! I combine them to dataset, convert it to dataframe and then sort_values. Finally, convert the dataframe back to dataset:

ds = cld.to_dataset(name='cld') ds['pair'] = pair

df = ds.to_dataframe() new_ds = df.sort_values(by='cld').to_xarray().transpose()

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/pydata/xarray/issues/3957#issuecomment-611299453, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIGDFO73HGCN7EOCCBW3NJDRLU2HHANCNFSM4MD6V32A .

You could access directly to data as ndarray and you could transform dataarray into a dataframe of pandas. Pandas has sort_values. You searched sorting values according z, it is shown in z index.

With more dataArray you could read about Dataset concept...

but i dont develop xarray, i am only user of that module, perhaps you search another type of answer.

http://xarray.pydata.org/en/stable/generated/xarray.Dataset.sortby.html according to values of 1-D dataarrays that share dimension with calling object.

El jue., 9 abr. 2020 4:22, Xin Zhang notifications@github.com escribió:

@JavierRuano https://github.com/JavierRuano Thank you very much. This example is a special case. If the order of z is different for each x and y, do we need to create a tmp DataArray to save the result of looping x and y ?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/pydata/xarray/issues/3957#issuecomment-611291129, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIGDFO4TP6RSAK7CV3DJV73RLUWPRANCNFSM4MD6V32A .

cld.reindex(z=cld[:,0,0].sortby(cld[:,0,0]).z)

with this solution [0] [1]

<xarray.DataArray (z: 5, y: 2, x: 4)> array([[[ 0. , 1. , 2. , 3. ], [ 4. , 5. , 6. , 7. ]],

   [[ 1. ,  2.5,  4. ,  5.5],
    [ 7. ,  8.5, 10. , 11.5]],

   [[ 8. ,  9. , 10. , 11. ],
    [12. , 13. , 14. , 15. ]],

   [[16. , 17. , 18. , 19. ],
    [20. , 21. , 22. , 23. ]],

   [[24. , 25. , 26. , 27. ],
    [28. , 29. , 30. , 31. ]]])

Coordinates: * z (z) int64 0 4 1 2 3 Dimensions without coordinates: y, x

[0] https://stackoverflow.com/questions/41077393/how-to-sort-the-index-of-a-xarray-dataset-dataarray

[1] https://github.com/pydata/xarray/issues/967

El mié., 8 abr. 2020 a las 14:06, Xin Zhang (notifications@github.com) escribió:

.sortby() only supports sorting DataArray by coords values. I'm trying to sort one DataArray (cld) by data values along one dim and sort another DataArray (pair) by the same order. MCVE Code Sample

import xarray as xrimport numpy as np

x = 4 y = 2 z = 4 data = np.arange(xyz).reshape(z, y, x)

3d array with coords

cld_1 = xr.DataArray(data, dims=['z', 'y', 'x'], coords={'z': np.arange(z)})

2d array without coords

cld_2 = xr.DataArray(np.arange(xy).reshape(y, x)1.5+1, dims=['y', 'x'])

expand 2d to 3d

cld_2 = cld_2.expand_dims(z=[4])

concat

cld = xr.concat([cld_1, cld_2], dim='z')

paired array

pair = cld.copy(data=np.arange(xy(z+1)).reshape(z+1, y, x)) print(cld)print(pair)

Output

<xarray.DataArray (z: 5, y: 2, x: 4)> array([[[ 0. , 1. , 2. , 3. ], [ 4. , 5. , 6. , 7. ]],

   [[ 8. ,  9. , 10. , 11. ],
    [12. , 13. , 14. , 15. ]],

   [[16. , 17. , 18. , 19. ],
    [20. , 21. , 22. , 23. ]],

   [[24. , 25. , 26. , 27. ],
    [28. , 29. , 30. , 31. ]],

   [[ 1. ,  2.5,  4. ,  5.5],
    [ 7. ,  8.5, 10. , 11.5]]])

Coordinates: * z (z) int64 0 1 2 3 4 Dimensions without coordinates: y, x

<xarray.DataArray (z: 5, y: 2, x: 4)> array([[[ 0, 1, 2, 3], [ 4, 5, 6, 7]],

   [[ 8,  9, 10, 11],
    [12, 13, 14, 15]],

   [[16, 17, 18, 19],
    [20, 21, 22, 23]],

   [[24, 25, 26, 27],
    [28, 29, 30, 31]],

   [[32, 33, 34, 35],
    [36, 37, 38, 39]]])

Coordinates: * z (z) int64 0 1 2 3 4 Dimensions without coordinates: y, x

Problem Description

I've tried argsort(): cld.argsort(axis=0), but the result is wrong:

<xarray.DataArray (z: 5, y: 2, x: 4)> array([[[0, 0, 0, 0], [0, 0, 0, 0]],

   [[4, 4, 4, 4],
    [4, 4, 4, 4]],

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

   [[2, 2, 2, 2],
    [2, 2, 2, 2]],

   [[3, 3, 3, 3],
    [3, 3, 3, 3]]], dtype=int64)

Coordinates: * z (z) int64 0 1 2 3 4 Dimensions without coordinates: y, x

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/pydata/xarray/issues/3957, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIGDFOYZYE2UCDJR4AAHAJLRLSAEZANCNFSM4MD6V32A .