id,node_id,number,title,user,state,locked,assignee,milestone,comments,created_at,updated_at,closed_at,author_association,active_lock_reason,draft,pull_request,body,reactions,performed_via_github_app,state_reason,repo,type 2101987013,PR_kwDOAMm_X85lJbZW,8672,Fix multiindex level serialization after reset_index,4160723,closed,0,,,6,2024-01-26T10:40:42Z,2024-02-23T01:22:17Z,2024-01-31T17:42:29Z,MEMBER,,0,pydata/xarray/pulls/8672," - [x] Closes #8628 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8672/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1864056633,PR_kwDOAMm_X85YovK-,8107,Better default behavior of the Coordinates constructor,4160723,closed,0,,,2,2023-08-23T21:42:51Z,2024-02-04T18:32:42Z,2023-08-31T07:35:47Z,MEMBER,,0,pydata/xarray/pulls/8107," - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` After working more on `Coordinates` I realize that the default behavior of its constructor could be more consistent with other Xarray objects. This PR changes this default behavior such that: - Pandas indexes are created for dimension coordinates if `indexes=None` (default). To create dimension coordinates with no index, just pass `indexes={}`. - If another `Coordinates` object is passed as input, its indexes are also added to the new created object. Since we don't support alignment / merge here, the following call raises an error: `xr.Coordinates(coords=xr.Coordinates(...), indexes={...})`. This PR introduces a breaking change since `Coordinates` are now exposed in v2023.8.0, which has just been released. It is a bit unfortunate but I think it may be OK for a fresh feature, especially if the next release will be soon after this one.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8107/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1879864306,PR_kwDOAMm_X85ZdmTF,8142,Dirty workaround for mypy 1.5 error,4160723,closed,0,,,8,2023-09-04T09:21:18Z,2023-09-07T16:04:55Z,2023-09-07T08:21:12Z,MEMBER,,0,pydata/xarray/pulls/8142,"I wanted to fix the following error with mypy 1.5: ``` xarray/core/dataset.py:505: error: Definition of ""__eq__"" in base class ""DatasetOpsMixin"" is incompatible with definition in base class ""Mapping"" [misc] ``` Which looks similar to https://github.com/python/mypy/issues/9319. It is weird that here it worked with mypy versions < 1.5, though. I don't know if there is a better fix, but I thought that redefining `__eq__` in `Dataset` would be a bit less dirty workaround than adding `type: ignore` in the class declaration. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8142/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1879652439,PR_kwDOAMm_X85Zc4ub,8141,Fix doctests: pandas 2.1 MultiIndex repr with nan,4160723,closed,0,,,0,2023-09-04T07:08:55Z,2023-09-05T08:35:37Z,2023-09-05T08:35:36Z,MEMBER,,0,pydata/xarray/pulls/8141,,"{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8141/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1862912829,PR_kwDOAMm_X85Yk15B,8102,Add `Coordinates.assign()` method,4160723,closed,0,,,0,2023-08-23T09:15:51Z,2023-09-01T13:28:16Z,2023-09-01T13:28:16Z,MEMBER,,0,pydata/xarray/pulls/8102," - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` - [x] New functions/methods are listed in `api.rst` This is consistent with the Dataset and DataArray `assign` methods (now that `Coordinates` is also exposed as public API). This allows writing: ```python midx = pd.MultiIndex.from_arrays([[""a"", ""a"", ""b"", ""b""], [0, 1, 0, 1]]) midx_coords = xr.Coordinates.from_pandas_multiindex(midx, ""x"") ds = xr.Dataset(coords=midx_coords.assign(y=[1, 2])) ``` which is quite common (at least in the tests) and a bit nicer than ```python ds = xr.Dataset(coords=midx_coords.merge({""y"": [1, 2]}).coords) ```","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8102/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 180638999,MDExOlB1bGxSZXF1ZXN0ODc3MTUzMDM=,1028,"Add `set_index`, `reset_index` and `reorder_levels` methods",4160723,closed,0,,,8,2016-10-03T13:22:24Z,2023-08-30T09:28:26Z,2016-12-27T17:03:00Z,MEMBER,,0,pydata/xarray/pulls/1028,"Another item in #719. I added tests and updated the docs, so this is ready for review. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1028/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1864650372,PR_kwDOAMm_X85YqtUk,8109,Better error message when trying to set an index from a scalar coordinate,4160723,closed,0,,,0,2023-08-24T08:18:13Z,2023-08-30T09:27:27Z,2023-08-30T07:13:15Z,MEMBER,,0,pydata/xarray/pulls/8109," - [x] Closes #4091 - [x] Tests added The message suggests using `.expand_dims()`.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8109/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 966983801,MDExOlB1bGxSZXF1ZXN0NzA5MTg3NDY2,5692,Explicit indexes,4160723,closed,0,,,46,2021-08-11T15:57:41Z,2023-08-30T09:26:37Z,2022-03-17T17:11:44Z,MEMBER,,0,pydata/xarray/pulls/5692," - [x] Closes many issues: - [x] closes #1366 - [x] closes #1408 - [x] closes #2489 - [x] closes #3432 - [x] closes #4542 - [x] closes #4955 - [x] closes #5202 - [x] closes #5645 - [x] closes #5691 - [x] closes #5697 - [x] closes #5700 - [x] closes #5727 - [x] closes #5953 - [x] closes #6183 - [x] closes #6313 - [x] Tests added - [x] Passes `pre-commit run --all-files` - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` - New functions/methods are listed in `api.rst` (new `Index` and `Indexes` API not public yet) Follow-up on #5636 (work in progress), supersedes #2195. This is likely to be going big, sorry in advance! It'll be safer to make a release before merging this PR. Current progress: - [x] create (default) indexes using the `Index` classes - [x] refactor default indexes created when 1st accessing `.xindexes` or `.indexes` - [x] support for non-default indexes (no public API yet) - [x] remove multi-index virtual coordinates (replace it by regular coordinates) - [x] refactor internal (text / html) formatting functions - [x] internal refactor of location-based selection (`.isel()`) - [x] internal refactor of label-based selection (`.sel()`) - [x] internal refactor of `.rename()` - Some changes in behavior (see comments below) - see #4108 - see #4107 - see #4417 - [x] internal refactor of `set_index` / `reset_index` - [x] internal refactor of `stack` / `unstack` - Some changes in behavior (see comments below) - [x] internal refactor of `Dataset.to_stacked_array` - [x] internal refactor of `swap_dims` - [x] internal refactor of `expand_dims` - [x] internal refactor of alignment - [x] internal refactor of `reindex` and `reindex_like` - [x] internal refactor of `interp` and `interp_like` - [x] internal refactor of merge - [x] internal refactor of concat - [x] internal refactor of computation - [x] internal refactor of copy - [x] internal refactor of `update`, `assign`, `__setitem__`, `del`, `drop_vars`, etc. - updates must not corrupt multi-coordinate indexes - [x] internal refactor of `set_coords` and `reset_coords` - internal refactor of `drop_sel` and `drop_isel` (maybe later) - [x] internal refactor of `pad` - [x] internal refactor of `shift` - [x] internal refactor of `roll` TODO: - [x] Uniformize Index API with Xarray's API - [x] rename `Index.query()` -> `Index.sel()`? - [x] rename `PandasMultiIndex.from_product()` -> `PandasMultiIndex.stack()`? Add `Index.stack()` and `Index.unstack()`. - [x] remove `Index.union()` and `Index.intersection()` - [x] Use `Index.create_variables()` internally - [x] remove `PandasIndex.from_pandas_index()` and `PandasMultiIndex.from_pandas_index()` (use constructor + `.create_variables()` instead) - [x] Review where `.xindexes` is used and use private API instead (`._indexes`) if possible for speed - [x] requires that `_indexes` always returns a mapping - [x] Use `from __future__ import annotations` in `indexes.py` - [x] Re-activate default indexes invariant check (with opt-out for some tests) In next PRs: - custom `Index.__repr__` and `Index._repr_inline_` - add an `Indexes` section in `DataArray` / `Dataset` reprs - update public API (`set_index`, `reset_index`, `drop_indexes`, `Dataset` and `DataArray` constructors, etc.) - allow multi-dimensional variables with `name` in `var.dims`","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5692/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 953235338,MDExOlB1bGxSZXF1ZXN0Njk3MzA3NDc3,5636,Refactor index vs. coordinate variable(s),4160723,closed,0,,,4,2021-07-26T19:54:25Z,2023-08-30T09:21:55Z,2021-08-09T07:56:56Z,MEMBER,,0,pydata/xarray/pulls/5636," - [x] Closes #5553 - [x] Tests added - [x] Passes `pre-commit run --all-files` - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` This implements option 3 (sort of) described in https://github.com/pydata/xarray/issues/5553#issue-933551030: - the goal is to avoid wrapping an `xarray.Index` into an `xarray.Variable` and keep those two concepts distinct from each other. - the `xarray.Index.from_variables` class constructor accepts a dictionary of `xarray.Variable` objects as argument and may (or should?) also return corresponding `xarray.IndexVariable` objects to ensure immutability. - for `PandasIndex`, the new returned `xarray.IndexVariable` wraps the underlying `pd.Index` via a `PandasIndexingAdapter` (this reverts some changes made in #5102). - for `PandasMultiIndex`, this PR adds `PandasMultiIndexingAdapter` so that we can wrap the pandas multi-index in separate coordinate variables objects: one for the dimension + one for each level. The level coordinates data internally hold a reference to the dimension coordinate data to avoid indexing the same underlying `pd.MultiIndex` for each of those coordinates (`PandasMultiIndexingAdapter.__getitem__` is memoized for that purpose). This is very much work in progress, I need to update (or revert) all related parts of Xarray's internals, update tests, etc. At this stage any comment on the approach described above is welcome. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5636/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1485037066,PR_kwDOAMm_X85Ez9Gj,7368,"Expose ""Coordinates"" as part of Xarray's public API",4160723,closed,0,,,31,2022-12-08T16:59:29Z,2023-08-30T09:11:57Z,2023-07-21T20:40:03Z,MEMBER,,0,pydata/xarray/pulls/7368," - [x] Closes #7214 - [x] Closes #6392 - [x] xref #6633 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` - [x] New functions/methods are listed in `api.rst` This is a rework of #7214. It follows the suggestions made in https://github.com/pydata/xarray/pull/7214#issuecomment-1295283938, https://github.com/pydata/xarray/pull/7214#issuecomment-1297046405 and https://github.com/pydata/xarray/pull/7214#issuecomment-1293774799: - No `indexes` argument is added to `Dataset.__init__`, and the `indexes` argument of `DataArray.__init__` is kept private (i.e., valid only if fastpath=True) - When a `Coordinates` object is passed to a new Dataset or DataArray via the `coords` argument, both coordinate variables and indexes are copied/extracted and added to the new object - This PR also adds ~~an `IndexedCoordinates` subclass~~ `Coordinates` public constructors used to create Xarray coordinates and indexes from non-Xarray objects. For example, the `Coordinates.from_pandas_multiindex()` class method creates a new set of index and coordinates from an existing `pd.MultiIndex`. EDIT: `IndexCoordinates` has been merged with `Coordinates` EDIT2: it ended up as a pretty big refactor with the promotion of `Coordinates` has a 2nd-class Xarray container that supports alignment like Dataset and DataArray. It is still quite advanced API, useful for passing coordinate variables and indexes around. Internally, `Coordinates` objects are still ""virtual"" containers (i.e., proxies for coordinate variables and indexes stored in their corresponding DataArray or Dataset objects). For now, a ""stand-alone"" `Coordinates` object created from scratch wraps a Dataset with no data variables. Some examples of usage: ```python import pandas as pd import xarray as xr midx = pd.MultiIndex.from_product([[""a"", ""b""], [1, 2]], names=(""one"", ""two"")) coords = xr.Coordinates.from_pandas_multiindex(midx, ""x"") # Coordinates: # * x (x) object MultiIndex # * one (x) object 'a' 'a' 'b' 'b' # * two (x) int64 1 2 1 2 ds = xr.Dataset(coords=coords) # # Dimensions: (x: 4) # Coordinates: # * x (x) object MultiIndex # * one (x) object 'a' 'a' 'b' 'b' # * two (x) int64 1 2 1 2 # Data variables: # *empty* ds_to_be_deprecated = xr.Dataset(coords={""x"": midx}) ds_to_be_deprecated.identical(ds) # True da = xr.DataArray([1, 2, 3, 4], dims=""x"", coords=ds.coords) # # array([1, 2, 3, 4]) # Coordinates: # * x (x) object MultiIndex # * one (x) object 'a' 'a' 'b' 'b' # * two (x) int64 1 2 1 2 ``` TODO: - [x] update `assign_coords` too so it has the same behavior if a `Coordinates` object is passed? - [x] How to avoid building any default index? It seems silly to add or use the `indexes` argument just for that purpose? ~~We could address that later.~~ Solution: wrap the coordinates dict in a Coordinates objects, e.g., `ds = xr.Dataset(coords=xr.Coordinates(coords_dict))`. @shoyer, @dcherian, anyone -- what do you think about the approach proposed here? I'd like to check that with you before going further with tests, docs, etc. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7368/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1422543378,PR_kwDOAMm_X85BgRaG,7214,Pass indexes directly to the DataArray and Dataset constructors,4160723,closed,0,,,17,2022-10-25T14:16:44Z,2023-08-30T09:11:56Z,2023-07-18T11:52:11Z,MEMBER,,1,pydata/xarray/pulls/7214," - [x] Closes #6392 - [x] Closes #6633 ? - [ ] Tests added - [ ] User visible changes (including notable bug fixes) are documented in `whats-new.rst` - [ ] New functions/methods are listed in `api.rst` From https://github.com/pydata/xarray/issues/6392#issuecomment-1290454937: I'm thinking of only accepting one or more instances of [Indexes](https://github.com/pydata/xarray/blob/e678a1d7884a3c24dba22d41b2eef5d7fe5258e7/xarray/core/indexes.py#L1030) as indexes argument in the Dataset and DataArray constructors. The only exception is when `fastpath=True` a mapping can be given directly. Also, when an empty collection of indexes is passed this skips the creation of default pandas indexes for dimension coordinates. - It is much easier to handle: just check that keys returned by `Indexes.variables` do no conflict with the coordinate names in the `coords` argument - It is slightly safer: it requires the user to explicitly create an `Indexes` object, thus with less chance to accidentally provide coordinate variables and index objects that do not relate to each other (we could probably add some safe guards in the `Indexes` class itself) - It is more convenient: an Xarray `Index` may provide a factory method that returns an instance of `Indexes` that we just need to pass as indexes, and we could also do something like `ds = xr.Dataset(indexes=other_ds.xindexes)` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7214/reactions"", ""total_count"": 1, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 1, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1863646946,PR_kwDOAMm_X85YnWau,8104,Fix merge with compat=minimal (coord names),4160723,closed,0,,,0,2023-08-23T16:20:48Z,2023-08-30T09:11:18Z,2023-08-30T07:57:35Z,MEMBER,,0,pydata/xarray/pulls/8104," - [x] Closes #7405 - [x] Closes #7588 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8104/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1358841264,PR_kwDOAMm_X84-NgIX,6975,Add documentation on custom indexes,4160723,closed,0,,,9,2022-09-01T13:20:00Z,2023-08-30T09:10:34Z,2023-07-17T23:23:22Z,MEMBER,,0,pydata/xarray/pulls/6975,"This PR documents the API of the `Index` base class and adds a guide for creating custom indexes (reworked from https://hackmd.io/Zxw_zCa7Rbynx_iJu6Y3LA). Hopefully it will help anyone experimenting with this feature. @pydata/xarray your feedback would be very much appreciated! I've been into this for quite some time, so there may be things that seem obvious to me but that you can still find very confusing or non-intuitive. It would then deserve some extra or better explanation. More specifically, I'm open to any suggestion on how to better illustrate this with clear and succinct examples. There are other parts of the documentation that still need to be updated regarding the indexes refactor (e.g., ""dimension"" coordinates, `xindexes` property, set/drop indexes, etc.). But I suggest to do that in separate PRs and focus here on creating custom indexes.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6975/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1859437888,PR_kwDOAMm_X85YY-II,8094,Refactor update coordinates to better handle multi-coordinate indexes,4160723,closed,0,,,4,2023-08-21T13:57:38Z,2023-08-30T09:06:28Z,2023-08-29T14:23:29Z,MEMBER,,0,pydata/xarray/pulls/8094," - [x] Closes #7563 - [x] Closes #8039 - [x] Closes #8056 - [x] Closes #7885 - [x] Closes #7921 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` This refactor should better handle multi-coordinate indexes when updating (or assigning) new coordinates. It also fixes, better isolates and better warns a bunch of deprecated pandas multi-index special cases (i.e., directly passing `pd.MultiIndex` objects or updating a multi-index dimension coordinate). I very much look forward to seeing support for those cases dropped :). ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/8094/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1498386428,PR_kwDOAMm_X85FiyaY,7382,Some alignment optimizations,4160723,closed,0,,,4,2022-12-15T12:54:56Z,2023-08-30T09:05:24Z,2023-01-05T21:25:55Z,MEMBER,,0,pydata/xarray/pulls/7382," - [x] Benchmark added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` May fix some performance regressions, e.g., see https://github.com/pydata/xarray/issues/7376#issuecomment-1352989233. @ravwojdyla with this PR `ds.assign(foo=~ds[""d3""])` in your example should be much faster (on par with version 2022.3.0).","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7382/reactions"", ""total_count"": 1, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 1, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1362148668,PR_kwDOAMm_X84-YVgW,6992,Review (re)set_index,4160723,closed,0,,,1,2022-09-05T15:07:43Z,2023-08-30T09:05:10Z,2022-09-27T10:35:38Z,MEMBER,,0,pydata/xarray/pulls/6992," - [x] Closes - [x] fixes #6946 - [x] fixes #6989 - [x] fixes #6959 - [x] fixes #6969 - [x] fixes #7036 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` Restore behavior prior to the explicit indexes refactor (i.e., refactored but without breaking changes). TODO: - [x] review `set_index` - [x] review `reset_index` For `reset_index`, the only behavior that is not restored here is the coordinate renamed with a `_` suffix when dropping a single index. This was originally to prevent any coordinate with no index matching a dimension name, which is now irrelevant. That is a quite dirty workaround and I don't know who is relying on it (no complaints yet), but I'm open to restore it if needed (esp. considering that we may later deprecate `reset_index` completely in favor of `drop_indexes` #6971).","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6992/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1307195361,PR_kwDOAMm_X847hz6o,6800,"(scipy 2022 branch) Add an ""options"" argument to Index.from_variables()",4160723,closed,0,,,1,2022-07-17T20:01:00Z,2022-12-08T09:38:50Z,2022-09-02T13:54:46Z,MEMBER,,0,pydata/xarray/pulls/6800,"It allows passing options to the constructor of a custom `Index` subclass, in case there's any relevant build options to expose to users. This could for example be the distance metric chosen for an index based on `sklearn.neighbors.BallTree`, or the CRS definition for a geospatial index. The `**options` arguments of `Dataset.set_xindex()` are passed through. An alternative way would be to pass options via coordinate metadata, like the `spatial_ref` coordinate in rioxarray. Perhaps both alternatives may co-exist? This PR also adds type annotations to `set_xindex()`. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6800/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1357296406,PR_kwDOAMm_X84-IR52,6971,Add set_xindex and drop_indexes methods,4160723,closed,0,,,7,2022-08-31T12:54:35Z,2022-12-08T09:38:13Z,2022-09-28T07:25:15Z,MEMBER,,0,pydata/xarray/pulls/6971," - [x] Closes #6849 - [x] Supersedes #6800 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` - [x] New functions/methods are listed in `api.rst` This PR adds Dataset and DataArray `.set_xindex` and `.drop_indexes` methods (the latter is also discussed in #4366). I've cherry picked the relevant commits in the `scipy22` branch and added a few more commits. This PR also allows passing build options to any `Index`. Some comments and open questions: - Should we make the `index_cls` argument of `set_xindex` optional? - I.e., `set_index(coord_names, index_cls=None, **options)` where a pandas index is created by default (or a pandas multi-index if several coordinate names are given), provided that the coordinate(s) are valid 1-d candidates. - This would be redundant with the existing `set_index` method, but this would be convenient if we later depreciate it. - Should we depreciate `set_index` and `reset_index`? I think we should, but probably not at this point yet. - There's a special case for multi-indexes where `set_xindex([""foo"", ""bar""], PandasMultiIndex)` adds a dimension coordinate in addition to the ""foo"" and ""bar"" level coordinates so that it is consistent with the rest of Xarray. I find it a bit annoying, though. Probably another motivation for depreciating this dimension coordinate. - In this PR I also imported the `Index` base class in Xarray's root namespace. - It is needed for custom indexes and it's just a little more convenient than importing it from `xarray.core.indexes`. - Should we do the same for `PandasIndex` and `PandasMultiIndex` subclasses? Maybe if one wants to create a custom index inheriting from it. `PandasMultiIndex` factory methods could be also useful if we depreciate passing `pd.MultiIndex` objects as DataArray / Dataset coordinates. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6971/reactions"", ""total_count"": 2, ""+1"": 2, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1363524666,PR_kwDOAMm_X84-c82D,6999,Raise UserWarning when rename creates a new dimension coord,4160723,closed,0,,,2,2022-09-06T16:16:17Z,2022-12-08T09:38:13Z,2022-09-27T09:33:40Z,MEMBER,,0,pydata/xarray/pulls/6999," - [x] Closes #6607 - [x] Closes #4107 - [x] Closes #6229 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` Current implemented ""fix"": raise a `UserWarning` and suggest using `swap_dims` (*) Alternatively, we could: - revert the breaking change (i.e., create the index again) and raise a `DeprecationWarning` instead - raise an error instead of a warning I don't have strong opinions on this, I'm happy to implement another alternative. The downside of reverting the breaking change now is that unfortunately it will introduce a breaking change in the next release., while workarounds are pretty straightforward. (*) from https://github.com/pydata/xarray/issues/6607#issuecomment-1126587818, doing `ds.set_coords(['lon']).rename(x='lon').set_index(lon='lon')` is working too. With #6971, `.set_xindex('lon')` could work as well. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6999/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1364493817,PR_kwDOAMm_X84-gJCw,7003,Misc. fixes for Indexes with pd.Index objects,4160723,closed,0,,,0,2022-09-07T11:05:02Z,2022-12-08T09:36:51Z,2022-09-23T07:30:38Z,MEMBER,,0,pydata/xarray/pulls/7003," - [x] Closes #6987 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7003/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1390999159,PR_kwDOAMm_X84_3QjW,7105,Fix to_index(): return multiindex level as single index,4160723,closed,0,,,4,2022-09-29T14:44:22Z,2022-12-08T09:36:51Z,2022-10-12T14:12:48Z,MEMBER,,0,pydata/xarray/pulls/7105," - [x] Closes #6836 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7105/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1193611401,PR_kwDOAMm_X841rm9D,6443,Fix concat with scalar coordinate (wrong index type),4160723,closed,0,,,1,2022-04-05T19:16:30Z,2022-12-08T09:36:50Z,2022-04-06T01:19:48Z,MEMBER,,0,pydata/xarray/pulls/6443," - [x] Closes #6434 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6443/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1389632629,PR_kwDOAMm_X84_ywy1,7101,Fix Dataset.assign_coords overwriting multi-index,4160723,closed,0,,,0,2022-09-28T16:21:48Z,2022-12-08T09:36:50Z,2022-09-28T18:02:16Z,MEMBER,,0,pydata/xarray/pulls/7101," - [x] Closes #7097 - [x] Tests added @dcherian the `DeprecationWarning` was ignored by default for `.assign_coords()` because of https://github.com/pydata/xarray/pull/6798#discussion_r924653224. I changed it to `FutureWarning` so that it is shown for both `.assign()` and `.assign_coords()`. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7101/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1324225268,PR_kwDOAMm_X848a7mk,6857,Fix aligned index variable metadata side effect,4160723,closed,0,,,0,2022-08-01T10:57:16Z,2022-12-08T09:36:49Z,2022-08-31T07:16:14Z,MEMBER,,0,pydata/xarray/pulls/6857," - [x] Closes #6852 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6857/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1472483025,PR_kwDOAMm_X85EHyv7,7347,Fix assign_coords resetting all dimension coords to default index,4160723,closed,0,,,3,2022-12-02T08:19:01Z,2022-12-08T09:36:49Z,2022-12-02T16:32:40Z,MEMBER,,0,pydata/xarray/pulls/7347," - [x] Closes #7346 - [x] Tests added - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/7347/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1176745736,PR_kwDOAMm_X840z4zt,6400,Speed-up multi-index html repr + add display_values_threshold option,4160723,closed,0,,,3,2022-03-22T12:57:37Z,2022-03-29T07:10:22Z,2022-03-29T07:05:32Z,MEMBER,,0,pydata/xarray/pulls/6400,"This adds `PandasMultiIndexingAdapter._repr_html_` that can greatly speed-up the html repr of Xarray objects with multi-indexes. This optimized `_repr_html_` implementation is now used for formatting the array detailed view of all multi-index coordinates in the html repr, instead of converting the full index and each levels to numpy arrays before formatting them. ```python import xarray as xr ds = xr.tutorial.load_dataset(""air_temperature"") da = ds[""air""].stack(z=[...]) da.shape # (3869000,) %timeit -n 1 -r 1 da._repr_html_() # 9.96 ms ! ``` - [x] Closes #5529 - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6400/reactions"", ""total_count"": 2, ""+1"": 2, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1174675456,PR_kwDOAMm_X840tJ9A,6388,isel: convert IndexVariable to Variable if index is dropped,4160723,closed,0,,,1,2022-03-20T20:29:58Z,2022-03-29T07:10:08Z,2022-03-21T04:47:48Z,MEMBER,,0,pydata/xarray/pulls/6388," - [x] Closes #6381 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6388/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 616432851,MDExOlB1bGxSZXF1ZXN0NDE2NTQ0MzE4,4053,Fix html repr in untrusted notebooks (plain text fallback),4160723,closed,0,,,5,2020-05-12T07:38:22Z,2022-03-29T07:10:07Z,2020-05-20T17:06:40Z,MEMBER,,0,pydata/xarray/pulls/4053," - [x] Closes #4041 - [x] Tests added - [x] Passes `isort -rc . && black . && mypy . && flake8` - [x] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API This is not very elegant (actually plain text repr is already included in the notebook as `text/plain` mime type but it is ignored when `text/html` mime type is present), but it seems to work. I haven't found a better workaround. I don't really know if this can be properly tested (I only added a basic test). Steps to test this fix: - To ""untrust"" a notebook: open an existing notebook with a simple editor, manually edit one output cell with a xarray object repr, and save the ipynb file. - Open this notebook with the Notebook app, you should see the plain text repr. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/4053/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 849315490,MDExOlB1bGxSZXF1ZXN0NjA4MTEwNjI0,5102,Flexible indexes: add Index base class and xindexes properties,4160723,closed,0,,,10,2021-04-02T16:18:07Z,2022-03-29T07:10:07Z,2021-05-11T08:21:26Z,MEMBER,,0,pydata/xarray/pulls/5102,"This PR clears up the path for flexible indexes: - it adds a new ~~`IndexAdapter`~~ `Index` base class that is meant to be inherited by all xarray-compatible indexes (built-in or 3rd-party) - `PandasIndexAdapter` now inherits from ~~`IndexAdapter`~~ `Index` - the `xarray_obj.xindexes` properties return `Index` (`PandasIndexAdapter`) instances. `xarray_obj.indexes` properties still return `pandas.Index` instances. ~~The latter is a breaking change, although I'm not sure if the `indexes` property has been made public yet.~~ This is still work in progress, there are many broken tests that are not fixed yet. (EDIT: all tests should be fixed now). There's a lot of dirty fixes to avoid circular dependencies and in the many places where we still need direct access to the `pandas.Index` objects, but I'd expect that these will be cleaned-up further in the refactoring.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5102/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 893415955,MDExOlB1bGxSZXF1ZXN0NjQ1OTMzODI3,5322,Internal refactor of label-based data selection,4160723,closed,0,,,1,2021-05-17T14:52:49Z,2022-03-29T07:10:07Z,2021-06-08T09:35:54Z,MEMBER,,0,pydata/xarray/pulls/5322,"Xarray label-based data selection now relies on a newly added `xarray.Index.query(self, labels: Dict[Hashable, Any]) -> Tuple[Any, Optional[None, Index]]` method where: - `labels` is a always a dictionary with coordinate name(s) as key(s) and the corresponding selection label(s) as values - When calling `.sel` with some coordinate(s)/label(s) pairs, those are first grouped by index so that only the relevant pairs are passed to an `Index.query` - the returned tuple contains the positional indexers and (optionally) a new index object For a simple `pd.Index`, `labels` always corresponds to a 1-item dictionary like `{'coord_name': label_values}`, which is not very useful in this case, but this format is useful for `pd.MultiIndex` and will likely be for other, custom indexes. Moving the label->positional indexer conversion logic into `PandasIndex.query()`, I've tried to separate `pd.Index` vs `pd.MultiIndex` concerns by adding a new `PandasMultiIndex` wrapper class (it will probably be useful for other things as well) and refactor the complex logic that was implemented in `convert_label_indexer`. Hopefully it is a bit clearer now. Working towards a more flexible/generic system, we still need to figure out how to: - pass index query extra arguments like `method` and `tolerance` for `pd.Index` but in a more generic way - handle several positional indexers over multiple dimensions possibly returned by a custom ""meta-index"" (e.g., staggered grid index) - handle the case of positional indexers returned from querying >1 indexes along the same dimension (e.g., multiple coordinates along `x` with a simple `pd.Index`) - pandas indexes don't need information like the names or shapes of their corresponding coordinate(s) to perform label-based selection, but this kind of information will probably be needed for other indexes (we actually need it for advanced point-wise selection using tree-based indexes in [xoak](https://github.com/xarray-contrib/xoak)). This could be done in follow-up PRs.. Side note: I've initially tried to return from `xindexes` items for multi-index levels as well (not only index dimensions), but it's probably wiser to save this for later (when we'll tackle the multi-index virtual coordinate refactoring) as there are many places in Xarray where this is clearly not expected. Happy to hear your thoughts @pydata/xarray.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5322/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 819062172,MDExOlB1bGxSZXF1ZXN0NTgyMjI0MTQ4,4979,Flexible indexes refactoring notes,4160723,closed,0,,,22,2021-03-01T16:57:32Z,2022-03-29T07:09:31Z,2021-03-17T16:47:29Z,MEMBER,,0,pydata/xarray/pulls/4979,"As a preliminary step before I take on the refactoring and implementation of flexible indexes in Xarray for the next few months, I reviewed the status of https://github.com/pydata/xarray/projects/1 and started compiling partially implemented or planned changes, thoughts, etc. into a single document that may serve as a basis for further discussion and implementation work. It's still very much work in progress (I will update it regularly in the forthcoming days) and it is very open to discussion (we can use this PR for that)! I'm not sure if Xarray's root folder is a good place for this document, though. We could move this into a new repository in `xarray-contrib` (that could also host other enhancement proposals) if that's necessary. I'm looking forward to getting started on this and to getting your thoughts/feedback! ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/4979/reactions"", ""total_count"": 13, ""+1"": 3, ""-1"": 0, ""laugh"": 0, ""hooray"": 7, ""confused"": 0, ""heart"": 3, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 903899735,MDExOlB1bGxSZXF1ZXN0NjU1MTA5NDg0,5385,Cast PandasIndex to pd.(Multi)Index,4160723,closed,0,,,0,2021-05-27T15:15:41Z,2022-03-29T07:09:31Z,2021-05-28T08:28:11Z,MEMBER,,0,pydata/xarray/pulls/5385," - [x] Closes #5384 - [x] Tests added - [x] Passes `pre-commit run --all-files` - [x] User visible changes (including notable bug fixes) are documented in `whats-new.rst` ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/5385/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1174687047,PR_kwDOAMm_X840tLrz,6389,Re-index: fix missing variable metadata,4160723,closed,0,,,2,2022-03-20T21:11:38Z,2022-03-29T07:09:31Z,2022-03-21T07:53:05Z,MEMBER,,0,pydata/xarray/pulls/6389," - [x] Closes #6382 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6389/reactions"", ""total_count"": 2, ""+1"": 2, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1174610081,PR_kwDOAMm_X840s_xU,6385,Fix concat with scalar coordinate,4160723,closed,0,,,0,2022-03-20T16:46:48Z,2022-03-29T07:09:30Z,2022-03-21T04:49:23Z,MEMBER,,0,pydata/xarray/pulls/6385," - [x] Closes #6384 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6385/reactions"", ""total_count"": 3, ""+1"": 3, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1174615799,PR_kwDOAMm_X840tAtL,6386,Fix Dataset groupby returning a DataArray,4160723,closed,0,,,0,2022-03-20T17:06:13Z,2022-03-29T07:09:30Z,2022-03-20T18:55:27Z,MEMBER,,0,pydata/xarray/pulls/6386," - [x] Closes #6379 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6386/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1175490214,PR_kwDOAMm_X840vt1_,6394,Fix DataArray groupby returning a Dataset,4160723,closed,0,,,0,2022-03-21T14:43:21Z,2022-03-29T07:09:30Z,2022-03-21T15:26:20Z,MEMBER,,0,pydata/xarray/pulls/6394," - [x] Closes #6393 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6394/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1174622308,PR_kwDOAMm_X840tBvD,6387,Fix concat with variable or dataarray as dim (propagate attrs),4160723,closed,0,,,1,2022-03-20T17:27:41Z,2022-03-29T07:09:29Z,2022-03-20T18:53:46Z,MEMBER,,0,pydata/xarray/pulls/6387," - [x] Closes #6380 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6387/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 1183360119,PR_kwDOAMm_X841JuRv,6418,Fix concat with scalar coordinate (dtype),4160723,closed,0,,,0,2022-03-28T12:22:50Z,2022-03-29T07:06:46Z,2022-03-28T16:05:01Z,MEMBER,,0,pydata/xarray/pulls/6418," - [x] Closes #6416 - [x] Tests added ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/6418/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 512564243,MDExOlB1bGxSZXF1ZXN0MzMyNTUyNTA3,3448,Add license for the icons used in the html repr,4160723,closed,0,,,1,2019-10-25T14:57:20Z,2019-10-25T15:48:52Z,2019-10-25T15:40:46Z,MEMBER,,0,pydata/xarray/pulls/3448,,"{""url"": ""https://api.github.com/repos/pydata/xarray/issues/3448/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 249584098,MDExOlB1bGxSZXF1ZXN0MTM1Mjk4ODY3,1507,Detailed report for testing.assert_equal and testing.assert_identical,4160723,closed,0,,,18,2017-08-11T09:38:23Z,2019-10-25T15:07:39Z,2019-01-18T09:16:31Z,MEMBER,,0,pydata/xarray/pulls/1507," - ~~Closes #xxxx~~ - [x] Tests added / passed - [x] Passes ``git diff upstream/master | flake8 --diff`` - [x] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API ~~In addition to `Dataset` repr, the error message also shows the output of `Dataset.info()` for both datasets.~~ ~~This may not be the most elegant solution, but it is helpful when datasets only differ by their attributes attached to coordinates or data variables (not shown in repr). I'm open to any suggestion.~~ The report shows the differences for dimensions, data values (``Variable`` and ``DataArray``), coordinates, data variables and attributes (the latter only for ``testing.assert_identical``). There is currently not much tests for `xarray.testing` functions, but I'm willing to add more if needed. Not sure if it's worth a what's new entry (EDIT: added one).","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1507/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 274619743,MDExOlB1bGxSZXF1ZXN0MTUzMTE4MjQ3,1723,Fix unexpected behavior of .set_index() since pandas 0.21.0,4160723,closed,0,,,0,2017-11-16T18:37:20Z,2019-10-25T15:07:18Z,2017-11-17T00:54:51Z,MEMBER,,0,pydata/xarray/pulls/1723," - [x] Closes #1722 - [x] Tests added / passed - [x] Passes ``git diff upstream/master **/*py | flake8 --diff`` - [x] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1723/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 287844110,MDExOlB1bGxSZXF1ZXN0MTYyNDI2NzU2,1820,WIP: html repr,4160723,closed,0,,,40,2018-01-11T16:33:07Z,2019-10-25T15:06:58Z,2019-10-24T16:48:46Z,MEMBER,,0,pydata/xarray/pulls/1820," - [x] Closes #1627 - [ ] Tests added - [ ] Tests passed - [ ] Passes ``git diff upstream/master **/*py | flake8 --diff`` - [ ] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API This is work in progress, although the basic functionality is there. You can see a preview here: http://nbviewer.jupyter.org/gist/benbovy/3009f342fb283bd0288125a1f7883ef2 TODO: - [ ] Add support for Multi-indexes - [ ] Probably good to have some opt-in or fail back system in case where we (or users) know that the rendering will not work - [ ] Add some tests Nice to have (keep this for later): - Clean-up CSS code and HTML template (track CSS [subgrid support](https://caniuse.com/#feat=css-subgrid) in browsers, this may simplify a lot the things here). - Dynamically adapt cell widths (given the length of the names of variables and dimensions). Currently all cells have a fixed width. This is tricky, though, as we don't use a monospace font here. - Integration with jupyterlab/notebook themes (CSS classes) and maybe allow custom CSS. - Integration of Dask arrays HTML repr (+ integration of repr for other array backends). - Maybe find a way (if possible) to include CSS only once in the notebook (currently it is included each time a xarray object is displayed in an output cell, which is not very nice). - Review the rules for collapsing the `Coordinates`, `Data variables` and `Attributes` sections (maybe expose them as global options). - Maybe also define some rules to collapse automatically the data section (DataArray and Variable) when the data repr is too long. - Maybe add rich representation for `Dataset.coords` and `Dataset.data_vars` as well?
Other thoughts (old) A big challenge here is to provide both robust and flexible styling (CSS): - I have tested the current styling in jupyterlab (0.30.6, light theme), notebook (5.2.2) and nbviewer: despite some slight differences it looks quite good! - However, the current CSS code is a bit fragile (I had to add a lot of `!important`). Probably this could be a bit cleaned and optimized (unfortunately my CSS skills are limited). - Also, with the jupyterlab's dark theme it looks ugly. We probably need to use jupyterlab CSS variables so that our CSS scheme is compatible with the theme machinery, but at the same time we need to support other front-ends. So we probably need to maintain different stylings (i.e., multiple CSS files, one of them picked-up depending on the front-end), though I don't know if it's easy to automatically detect the front-end (choosing a default style is difficult too). - The notebook rendering on Github seems to disable style tags (no style is applied to the output, see https://gist.github.com/benbovy/3009f342fb283bd0288125a1f7883ef2). Output is not readable at all in this case, so it might be useful to allow turning off rich output as an option.
","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1820/reactions"", ""total_count"": 3, ""+1"": 3, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 349078381,MDExOlB1bGxSZXF1ZXN0MjA3Mjc3NDg2,2357,DOC: move xarray related projects to top-level TOC section,4160723,closed,0,,,1,2018-08-09T10:57:47Z,2018-08-11T13:41:24Z,2018-08-10T20:13:08Z,MEMBER,,0,pydata/xarray/pulls/2357,"Make xarray-related projects more discoverable, as it has been suggested in xarray mailing-list. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/2357/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 300588788,MDExOlB1bGxSZXF1ZXN0MTcxNjMxNTQ1,1946,DOC: add main sections to toc,4160723,closed,0,,,0,2018-02-27T11:13:17Z,2018-02-27T21:16:18Z,2018-02-27T19:04:24Z,MEMBER,,0,pydata/xarray/pulls/1946,"Not a big change, but adds a little more clarity IMO. I'm open to any suggestion for better section names and/or organization. Also I let ""What's new"" at the top, but not sure if ""Getting started"" is the right section.","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1946/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 230631480,MDExOlB1bGxSZXF1ZXN0MTIxOTQyNjMx,1422,xarray.core.variable.as_variable part of the public API,4160723,closed,0,,,6,2017-05-23T08:44:08Z,2017-06-10T18:33:34Z,2017-06-02T17:55:12Z,MEMBER,,0,pydata/xarray/pulls/1422," - [x] Closes #1303 - [x] Tests added / passed - [x] Passes ``git diff upstream/master | flake8 --diff`` (if we ignore messages for .rst files and ""imported but not used"" messages for `xarray.__init__.py`) - [x] Fully documented, including `whats-new.rst` for all changes and `api.rst` for new API Make `xarray.core.variable.as_variable` part of the public API and accessible as a top-level function: `xarray.as_variable`. I changed the docstrings to follow the numpydoc format more closely. I also removed the `copy=False` keyword arguments as apparently it was unused. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/1422/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 169588316,MDExOlB1bGxSZXF1ZXN0ODAyMjk0OTM=,947,Multi-index levels as coordinates,4160723,closed,0,,,17,2016-08-05T11:34:49Z,2016-09-14T03:35:04Z,2016-09-14T03:34:51Z,MEMBER,,0,pydata/xarray/pulls/947,"Implements 2, 4 and 5 in #719. Demo: ``` In [1]: import numpy as np In [2]: import pandas as pd In [3]: import xarray as xr In [4]: index = pd.MultiIndex.from_product((list('ab'), range(2)), ...: names= ('level_1', 'level_2')) In [5]: da = xr.DataArray(np.random.rand(4, 4), coords={'x': index}, ...: dims=('x', 'y'), name='test') In [6]: da Out[6]: array([[ 0.15036153, 0.68974802, 0.40082234, 0.94451318], [ 0.26732938, 0.49598123, 0.8679231 , 0.6149102 ], [ 0.3313594 , 0.93857424, 0.73023367, 0.44069622], [ 0.81304837, 0.81244159, 0.37274953, 0.86405196]]) Coordinates: * level_1 (x) object 'a' 'a' 'b' 'b' * level_2 (x) int64 0 1 0 1 * y (y) int64 0 1 2 3 In [7]: da['level_1'] Out[7]: array(['a', 'a', 'b', 'b'], dtype=object) Coordinates: * level_1 (x) object 'a' 'a' 'b' 'b' * level_2 (x) int64 0 1 0 1 In [8]: da.sel(x='a', level_2=1) Out[8]: array([ 0.26732938, 0.49598123, 0.8679231 , 0.6149102 ]) Coordinates: x object ('a', 1) * y (y) int64 0 1 2 3 In [9]: da.sel(level_2=1) Out[9]: array([[ 0.26732938, 0.49598123, 0.8679231 , 0.6149102 ], [ 0.81304837, 0.81244159, 0.37274953, 0.86405196]]) Coordinates: * level_1 (level_1) object 'a' 'b' * y (y) int64 0 1 2 3 ``` Some notes about the implementation: - I slightly modified `Coordinate` so that it allows setting different values for the names of the coordinate and its dimension. There is no breaking change. - I also added a `Coordinate.get_level_coords` method to get independent, single-index coordinates objects from a MultiIndex coordinate. Remaining issues: - `Coordinate.get_level_coords` calls `pandas.MultiIndex.get_level_values` for each level and is itself called each time when indexing and for repr. This can be very costly!! It would be nice to return some kind of lazy index object instead of computing the actual level values. - repr replace a MultiIndex coordinate by its level coordinates. That can be confusing in some cases (see below). Maybe we can set a different marker than `*` for level coordinates. ``` In [6]: [name for name in da.coords] Out[6]: ['x', 'y'] In [7]: da.coords.keys() Out[7]: KeysView(Coordinates: * level_1 (x) object 'a' 'a' 'b' 'b' * level_2 (x) int64 0 1 0 1 * y (y) int64 0 1 2 3) ``` - `DataArray.level_1` doesn't return another `DataArray` object: ``` In [10]: da.level_1 Out[10]: array(['a', 'a', 'b', 'b'], dtype=object) ``` - Maybe we need to test the uniqueness of level names at `DataArray` or `Dataset` creation. Of course still needs proper tests and docs... ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/947/reactions"", ""total_count"": 1, ""+1"": 1, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 159768214,MDExOlB1bGxSZXF1ZXN0NzM0NjU0MTA=,879,Multi-index repr,4160723,closed,0,,,2,2016-06-11T10:58:13Z,2016-08-31T21:40:59Z,2016-08-31T21:40:59Z,MEMBER,,0,pydata/xarray/pulls/879,"Another item of #719. An example: ``` python >>> index = pd.MultiIndex.from_product((list('ab'), range(10))) >>> index.names= ('a_long_level_name', 'level_1') >>> data = xr.DataArray(range(20), [('x', index)]) >>> data array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]) Coordinates: * x (x) object MultiIndex - a_long_level_name object 'a' 'a' 'a' 'a' 'a' 'a' 'a' 'a' 'a' 'a' 'b' ... - level_1 int64 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 ``` To be consistent with the displayed coordinates and/or data variables, it displays the actual used level values. Using the `pandas.MultiIndex.get_level_values` method would be expensive for big indexes, so I re-implemented it in xarray so that we can truncate the computation to the first _x_ values, which is very cheap. It still needs testing. Maybe it would be nice to align the level values. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/879/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 166299782,MDExOlB1bGxSZXF1ZXN0Nzc5NTM1MjI=,903,fixed multi-index copy test,4160723,closed,0,,,1,2016-07-19T10:37:36Z,2016-07-19T14:48:12Z,2016-07-19T14:47:58Z,MEMBER,,0,pydata/xarray/pulls/903,,"{""url"": ""https://api.github.com/repos/pydata/xarray/issues/903/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 143264649,MDExOlB1bGxSZXF1ZXN0NjQwNDI5ODk=,802,Multi-index indexing,4160723,closed,0,,,22,2016-03-24T14:39:38Z,2016-07-19T10:48:56Z,2016-07-19T01:15:42Z,MEMBER,,0,pydata/xarray/pulls/802,"Follows #767. This is incomplete (it still needs some tests and documentation updates), but it is working for both `Dataset` and `DataArray` objects. I also don't know if it is fully compatible with lazy indexing (Dask). Using the example from #767: ``` In [4]: da.sel(band_wavenumber={'band': 'foo'}) Out[4]: array([ 0.00017, 0.00014]) Coordinates: * wavenumber (wavenumber) float64 4.05e+03 4.05e+03 ``` As shown in this example, similarily to pandas, it automatically renames the dimension and assigns a new coordinate when the selection doesn't return a `pd.MultiIndex` (here it returns a `pd.FloatIndex`). In some cases this behavior may be unwanted (??), so I added a `drop_level` keyword argument (if `False` it keeps the multi-index and doesn't change the dimension/coordinate names): ``` In [5]: da.sel(band_wavenumber={'band': 'foo'}, drop_level=False) Out[5]: array([ 0.00017, 0.00014]) Coordinates: * band_wavenumber (band_wavenumber) object ('foo', 4050.2) ('foo', 4050.3) ``` Note that it also works with `DataArray.loc`, but (for now) in that case it always returns the multi-index: ``` In [6]: da.loc[{'band_wavenumber': {'band': 'foo'}}] Out[6]: array([ 0.00017, 0.00014]) Coordinates: * band_wavenumber (band_wavenumber) object ('foo', 4050.2) ('foo', 4050.3) ``` This is however inconsistent with `Dataset.sel` and `Dataset.loc` that both apply `drop_level=True` by default, due to their different implementation. Two solutions: (1) make `DataArray.loc` apply drop_level by default, or (2) use `drop_level=False` by default everywhere. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/802/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull 159920667,MDExOlB1bGxSZXF1ZXN0NzM1NTQ2MTI=,881,Fix variable copy with multi-index ,4160723,closed,0,,,1,2016-06-13T10:38:46Z,2016-06-16T21:01:11Z,2016-06-16T21:01:07Z,MEMBER,,0,pydata/xarray/pulls/881,"Fixes #769. ","{""url"": ""https://api.github.com/repos/pydata/xarray/issues/881/reactions"", ""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,,13221727,pull