You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+12-4Lines changed: 12 additions & 4 deletions
Original file line number
Diff line number
Diff line change
@@ -11,11 +11,17 @@ Interface for using [cubed](https://github.com/tomwhite/cubed) with [xarray](htt
11
11
12
12
## Installation
13
13
14
-
Install via pip.
14
+
Install via pip
15
+
16
+
`pip install cubed-xarray`
17
+
18
+
or conda
19
+
20
+
`conda install -c conda-forge cubed-xarray`
15
21
16
22
## Importing
17
23
18
-
You don't need to import this package in user code. Once `pip install`-ed, xarray should automatically become aware of this package via the magic of entrypoints.
24
+
You don't need to import this package in user code. Once poperly installed, xarray should automatically become aware of this package via the magic of entrypoints.
19
25
20
26
## Usage
21
27
@@ -26,9 +32,9 @@ Xarray objects backed by cubed arrays can be created either by:
26
32
3. Passing a `chunks` argument to `xarray.open_dataset`.
27
33
28
34
In (2) and (3) the choice to use `cubed.Array` instead of `dask.array.Array` is made by passing the keyword argument `chunked_array_type='cubed'`.
29
-
To pass arguments to the constructor of `cubed.Array`then pass them via the dictionary `from_array_kwargs`, e.g. `from_array_kwargs={'spec': cubed.Spec(max_mem=2_000_000)}`.
35
+
To pass arguments to the constructor of `cubed.Array`you should pass them via the dictionary `from_array_kwargs`, e.g. `from_array_kwargs={'spec': cubed.Spec(allowed_mem='2GB')}`.
30
36
31
-
If cubed and cubed-xarray are installed but dask is not, then specifying the parallel array type to use is not necessary,
37
+
If cubed and cubed-xarray are installed but dask is not, then specifying `chunked_array_type` is not necessary,
32
38
as the entrypoints system will then default to the only chunked parallel backend available (i.e. cubed).
33
39
34
40
## Sharp Edges 🔪
@@ -42,6 +48,8 @@ and some other things _might_ work but have not yet been tried:
42
48
- Groupby
43
49
- Saving to formats other than zarr
44
50
51
+
In general a bug could take the form of an error, or of a silent attempt to coerce the array type to numpy by immediately computing the underlying array.
52
+
45
53
## Tests
46
54
47
55
Integration tests for wrapping cubed with xarray also live in this repository.
0 commit comments