GroupBy(multiple variables)

Author: dcherian

src/posts/multiple-groupers/index.md

@@ -0,0 +1,125 @@
+---
+title: 'Grouping by multiple arrays with Xarray'
+date: '2023-07-18'
+authors:
+  - name: Deepak Cherian
+    github: dcherian
+
+summary: 'Xarray finally supports grouping by multiple arrays. 🎉'
+---
+
+## TLDR
+
+Xarray now supports grouping by multiple variables ([docs](https://docs.xarray.dev/en/latest/user-guide/groupby.html#grouping-by-multiple-variables)). 🎉 😱 🤯 🥳. Try it out!
+
+## How do I use it?
+
+Install `xarray>=2024.08.0` and optionally [flox](https://flox.readthedocs.io/en/latest/) for better performance with reductions.
+
+## Simple example
+
+Set up a multiple variable groupby using [Grouper objects](https://docs.xarray.dev/en/latest/user-guide/groupby.html#grouping-by-multiple-variables).
+
+```python
+import xarray as xr
+from xarray.groupers import UniqueGrouper
+
+da = xr.DataArray(
+    np.array([1, 2, 3, 0, 2, np.nan]),
+    dims="d",
+    coords=dict(
+        labels1=("d", np.array(["a", "b", "c", "c", "b", "a"])),
+        labels2=("d", np.array(["x", "y", "z", "z", "y", "x"])),
+    ),
+)
+
+gb = da.groupby(labels1=UniqueGrouper(), labels2=UniqueGrouper())
+gb
+```
+
+```
+<DataArrayGroupBy, grouped over 2 grouper(s), 9 groups in total:
+	'labels1': 3 groups with labels 'a', 'b', 'c'
+	'labels2': 3 groups with labels 'x', 'y', 'z'>
+```
+
+Reductions work as usual:
+
+```python
+gb.mean()
+```
+
+```
+xarray.DataArray (labels1: 3, labels2: 3)> Size: 72B
+array([[1. , nan, nan],
+       [nan, 2. , nan],
+       [nan, nan, 1.5]])
+Coordinates:
+  * labels1  (labels1) object 24B 'a' 'b' 'c'
+  * labels2  (labels2) object 24B 'x' 'y' 'z'
+```
+
+So does `map`:
+
+```python
+gb.map(lambda x: x[0])
+```
+
+```
+<xarray.DataArray (labels1: 3, labels2: 3)> Size: 72B
+array([[ 1., nan, nan],
+       [nan,  2., nan],
+       [nan, nan,  3.]])
+Coordinates:
+  * labels1  (labels1) object 24B 'a' 'b' 'c'
+  * labels2  (labels2) object 24B 'x' 'y' 'z'
+```
+
+## Multiple Groupers
+
+Combining different grouper types is allowed, that is you can combine
+categorical grouping with` UniqueGrouper`, binning with `BinGrouper`, and
+resampling with `TimeResampler`.
+
+```python
+ds = xr.Dataset(
+        {"foo": (("x", "y"), np.arange(12).reshape((4, 3)))},
+        coords={"x": [10, 20, 30, 40], "letters": ("x", list("abba"))},
+    )
+gb = ds.groupby(x=BinGrouper(bins=[5, 15, 25]), letters=UniqueGrouper())
+gb
+```
+
+```
+from xarray.groupers import BinGrouper
+
+ds = xr.Dataset(
+        {"foo": (("x", "y"), np.arange(12).reshape((4, 3)))},
+        coords={"x": [10, 20, 30, 40], "letters": ("x", list("abba"))},
+    )
+gb = ds.foo.groupby(x=BinGrouper(bins=[5, 15, 25]), letters=UniqueGrouper())
+gb
+```
+
+```
+<DatasetGroupBy, grouped over 2 grouper(s), 4 groups in total:
+	'x_bins': 2 groups with labels (5,, 15], (15,, 25]
+	'letters': 2 groups with labels 'a', 'b'>
+```
+
+```python
+gb.mean()
+```
+
+```
+<xarray.DataArray 'foo' (x_bins: 2, letters: 2, y: 3)> Size: 96B
+array([[[ 0.,  1.,  2.],
+        [nan, nan, nan]],
+
+       [[nan, nan, nan],
+        [ 3.,  4.,  5.]]])
+Coordinates:
+  * x_bins   (x_bins) object 16B (5, 15] (15, 25]
+  * letters  (letters) object 16B 'a' 'b'
+Dimensions without coordinates: y
+```