使用 xarray 为变量分配新坐标

Question

我有一个 .nc 文件作为 xarray 中的数据集打开，其结构如下：

ds

<xarray.Dataset>
Dimensions:    (lat: 733, lon: 720, time: 204)
Coordinates:
  * time       (time) int16 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ...
Dimensions without coordinates: lat , lon
Data variables :
    latitude   (lat) float32 56.0 55.9917 55.9833 55.975 55.9667 55.9583 ...
    longitude  (lon) float32 -11.0 -10.9917 -10.9833 -10.975 -10.9667 ...
    n2o        (lat, lon, time) float64 nan nan nan nan nan nan nan nan nan ...
    co2        (lat, lon, time) float64 nan nan nan nan nan nan nan nan nan ...
    no3        (lat, lon, time) float64 nan nan nan nan nan nan nan nan nan ...
    ch4        (lat, lon, time) float64 nan nan nan nan nan nan nan nan nan ...
    soc        (lat, lon, time) float64 nan nan nan nan nan nan nan nan nan ...

此文件存在问题，变量（n2o、co2 等）的坐标不正确，与它们关联的纬度和经度错误。我想将“纬度”和“经度”数据变量（包含正确值）分配为其他变量（n2o、co2 等）的坐标，以替换“纬度”和“经度”。我不确定这是否可能。

“纬度”变量如下所示：

<xarray.Variable (lat: 733)>
array([ 56.      ,  55.991665,  55.98333 , ...,  49.915367,  49.907032,
    49.898697], dtype=float32)
Attributes:
    units:      degrees of latitude North to South in 30 arc seconds steps
    long_name:  latitude

该ds.co2.lat尺寸看起来是这样的：

<xarray.DataArray 'lat' (lat: 733)>
array([  0,   1,   2, ..., 730, 731, 732], dtype=int64)
Dimensions without coordinates: lat

我试过了：

newlat = ds.variables['latitude'][:]

ds.co2['latitude'] = newlat

这似乎没有做任何事情 - 我不知道接下来要做什么。

Answer 1

您可能正在寻找rename和set_coords方法。您所看到的是 xarray 无法确定纬度/经度是坐标变量。解决方案是首先重命名纬度/经度变量，然后将这些变量关联为坐标。

In [10]: ds = xr.Dataset({'a': xr.Variable(('lon', 'lat'), np.random.random((4, 5))), 'b': xr.Variable(('lon', 'lat'),np.random.random((4, 5))), 'longitude':
    ...:  xr.Variable('lon', [1, 2, 3, 4]), 'latitude': xr.Variable('lat', [1, 2, 3, 4, 5])})

In [11]: ds
Out[11]:
<xarray.Dataset>
Dimensions:    (lat: 5, lon: 4)
Dimensions without coordinates: lat, lon
Data variables:
    a          (lon, lat) float64 0.8694 0.5929 0.04661 0.9571 0.1814 ...
    b          (lon, lat) float64 0.8893 0.5772 0.8457 0.8337 0.7966 0.1619 ...
    longitude  (lon) int64 1 2 3 4
    latitude   (lat) int64 1 2 3 4 5

In [12]: ds.rename({'latitude': 'lat', 'longitude': 'lon'}).set_coords(['lon', 'lat'])
Out[12]:
<xarray.Dataset>
Dimensions:  (lat: 5, lon: 4)
Coordinates:
  * lon      (lon) int64 1 2 3 4
  * lat      (lat) int64 1 2 3 4 5
Data variables:
    a        (lon, lat) float64 0.8694 0.5929 0.04661 0.9571 0.1814 0.01788 ...
    b        (lon, lat) float64 0.8893 0.5772 0.8457 0.8337 0.7966 0.1619 ...