xarray指南：索引和选择数据 - 最近邻查找
目录
本文翻译自 xarray 官方文档 Indexing and selecting data 的部分内容。
本文使用之前文章《xarray指南：索引和选择数据 - 位置和名称索引》中的示例数据，不再重复介绍。
本文介绍如何使用 xarray 实现最近邻查找。
介绍
基于标签的选择方法 sel()，reindex() 和 reindex_like() 都支持 method 和 tolerance 关键字参数。
method 参数允许使用 pad，backfill 或 nearest 方法启用最近邻（inexact）查找：
构造一个一维数组
da = xr.DataArray(
[1, 2, 3],
[("x", [0, 1, 2])]
)
da<xarray.DataArray (x: 3)>
array([1, 2, 3])
Coordinates:
* x (x) int64 0 1 2
选择给定坐标的最近邻点
da.sel(x=[1.1, 1.9], method="nearest")<xarray.DataArray (x: 2)>
array([2, 3])
Coordinates:
* x (x) int64 1 2
使用 backfill 选择下一个有效值
da.sel(x=0.1, method="backfill")<xarray.DataArray ()>
array(2)
Coordinates:
x int64 1
按照给定的坐标重新构造数组。
da.reindex(
x=[0.5, 1, 1.5, 2, 2,5],
method="pad",
)<xarray.DataArray (x: 6)>
array([1, 2, 2, 3, 3, 3])
Coordinates:
* x (x) float64 0.5 1.0 1.5 2.0 2.0 5.0
tolerance 限制了使用不精确查找进行有效匹配的最大距离：
da.reindex(
x=[1.1, 1.5],
method="nearest",
tolerance=0.2,
)<xarray.DataArray (x: 2)>
array([ 2., nan])
Coordinates:
* x (x) float64 1.1 1.5
如果 .sel() 的任何参数是切片对象，则尚不支持 method 参数：
# 会抛出异常
da.sel(x=slice(1, 3), method="nearest")但是，您无需使用 method 来进行不精确的切片。
只要索引标签是单调递增的，切片就已经返回范围内（包括）的所有值：
da.sel(x=slice(0.9, 3.1))<xarray.DataArray (x: 2)>
array([2, 3])
Coordinates:
* x (x) int64 1 2
只要 slice 或 .loc 参数也减小，带有单调递减标签的索引轴也可以工作：
reversed_da = da[::-1]
reversed_da.loc[3.1:0.9]<xarray.DataArray (x: 2)>
array([3, 2])
Coordinates:
* x (x) int64 2 1
如果要沿坐标插值而不是查找最近邻，请使用 interp() 和 interp_like()。
实战
t850.sel(latitude=[80, 90], method="nearest")<xarray.DataArray 't' (latitude: 2, longitude: 1440)>
array([[263.32236, 263.37234, 263.43234, ..., 263.20233, 263.25235, 263.28235],
[249.19234, 249.16234, 249.16234, ..., 249.15234, 249.19234, 249.14235]],
dtype=float32)
Coordinates:
time datetime64[ns] ...
step timedelta64[ns] ...
isobaricInhPa int64 ...
* latitude (latitude) float64 80.12 89.88
* longitude (longitude) float64 0.0 0.25 0.5 0.75 ... 359.2 359.5 359.8
valid_time datetime64[ns] ...
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
t850.sel(latitude=50, method="backfill")<xarray.DataArray 't' (longitude: 1440)>
array([275.05234, 275.04236, 275.18234, ..., 275.81235, 275.46234, 275.19235],
dtype=float32)
Coordinates:
time datetime64[ns] ...
step timedelta64[ns] ...
isobaricInhPa int64 ...
latitude float64 49.88
* longitude (longitude) float64 0.0 0.25 0.5 0.75 ... 359.2 359.5 359.8
valid_time datetime64[ns] ...
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
t850.reindex(latitude=[10, 20, 30, 40, 50], method="pad")t850.reindex(latitude=[10, 20, 30, 40, 50], method="pad")
t850.reindex(latitude=[10, 20, 30, 40, 50], method="pad")
<xarray.DataArray 't' (latitude: 5, longitude: 1440)>
array([[295.08234, 295.09235, 295.07236, ..., 295.05234, 295.03235, 295.04236],
[288.99234, 289.02234, 289.05234, ..., 288.86234, 288.91235, 288.95233],
[280.86234, 281.12234, 281.23233, ..., 279.51233, 279.93234, 280.38235],
[276.71234, 276.92233, 276.90234, ..., 276.45233, 276.51233, 276.61234],
[274.84235, 274.85236, 274.94235, ..., 275.43234, 275.14233, 274.94235]],
dtype=float32)
Coordinates:
* latitude (latitude) int64 10 20 30 40 50
time datetime64[ns] 2020-03-18
step timedelta64[ns] 4 days 09:00:00
isobaricInhPa int64 850
* longitude (longitude) float64 0.0 0.25 0.5 0.75 ... 359.2 359.5 359.8
valid_time datetime64[ns] 2020-03-22T09:00:00
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
t850.reindex(
latitude=[10, 10.2, 10.5],
method="nearest",
tolerance=0.1
)<xarray.DataArray 't' (latitude: 3, longitude: 1440)>
array([[ nan, nan, nan, ..., nan, nan,
nan],
[295.08234, 295.09235, 295.07236, ..., 295.05234, 295.03235,
295.04236],
[ nan, nan, nan, ..., nan, nan,
nan]], dtype=float32)
Coordinates:
* latitude (latitude) float64 10.0 10.2 10.5
time datetime64[ns] 2020-03-18
step timedelta64[ns] 4 days 09:00:00
isobaricInhPa int64 850
* longitude (longitude) float64 0.0 0.25 0.5 0.75 ... 359.2 359.5 359.8
valid_time datetime64[ns] 2020-03-22T09:00:00
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
latitude 已经按照从大到小排序。
t850.sel(latitude=slice(30, 20))<xarray.DataArray 't' (latitude: 40, longitude: 1440)>
array([[281.30234, 281.41235, 281.63235, ..., 280.18234, 280.52234, 280.92233],
[281.55234, 281.72235, 282.12234, ..., 280.71234, 280.96234, 281.31235],
[281.73233, 282.06235, 282.68234, ..., 281.19235, 281.37234, 281.56235],
...,
[288.52234, 288.58234, 288.62234, ..., 288.38235, 288.42233, 288.47235],
[288.74234, 288.78235, 288.81235, ..., 288.60236, 288.65234, 288.70233],
[288.99234, 289.02234, 289.05234, ..., 288.86234, 288.91235, 288.95233]],
dtype=float32)
Coordinates:
time datetime64[ns] ...
step timedelta64[ns] ...
isobaricInhPa int64 ...
* latitude (latitude) float64 29.88 29.62 29.38 ... 20.62 20.38 20.12
* longitude (longitude) float64 0.0 0.25 0.5 0.75 ... 359.2 359.5 359.8
valid_time datetime64[ns] ...
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
检索北京基本站（54511）最近邻的温度数据。
t850.sel(
latitude=39.48,
longitude=116.28,
method="nearest",
)<xarray.DataArray 't' ()>
array(281.94235, dtype=float32)
Coordinates:
time datetime64[ns] ...
step timedelta64[ns] ...
isobaricInhPa int64 ...
latitude float64 39.38
longitude float64 116.2
valid_time datetime64[ns] ...
Attributes:
GRIB_paramId: 130
...skip...
standard_name: air_temperature
参考
http://xarray.pydata.org/en/stable/indexing.html#nearest-neighbor-lookups
