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
Accelerated and Python-only metrics for benchmarking single-cell integration outputs.
11
30
12
-
This package contains implementations of metrics for evaluating the performance of single-cell omics data integration methods. The implementations of these metrics use [jax](https://jax.readthedocs.io/en/latest/) when possible for jit-compilation and hardware acceleration. All implementations are in Python.
31
+
This package contains implementations of metrics for evaluating the performance of single-cell omics data integration methods. The implementations of these metrics use [JAX](https://jax.readthedocs.io/en/latest/) when possible for jit-compilation and hardware acceleration. All implementations are in Python.
13
32
14
33
Currently we are porting metrics used in the scIB [manuscript](https://www.nature.com/articles/s41592-021-01336-8) (and [code](https://github.com/theislab/scib)). Deviations from the original implementations are documented. However, metric values from this repository should not be compared to the scIB repository.
15
34
@@ -22,15 +41,7 @@ Please refer to the [documentation][link-docs].
22
41
You need to have Python 3.9 or newer installed on your system. If you don't have
23
42
Python installed, we recommend installing [Miniconda](https://docs.conda.io/en/latest/miniconda.html).
24
43
25
-
There are several alternative options to install scib-metrics:
26
-
27
-
<!--
28
-
1) Install the latest release of `scib-metrics` from `PyPI <https://pypi.org/project/scib-metrics/>`_:
29
-
30
-
```bash
31
-
pip install scib-metrics
32
-
```
33
-
-->
44
+
There are several options to install scib-metrics:
To leverage hardward acceleration (e.g., GPU) please install the apprpriate version of [jax](https://github.com/google/jax#installation) separately. Often this can be easier by using conda-distributed versions of jax.
58
+
To leverage hardware acceleration (e.g., GPU) please install the apprpriate version of [JAX](https://github.com/google/jax#installation) separately. Often this can be easier by using conda-distributed versions of JAX.
48
59
49
60
## Release notes
50
61
51
62
See the [changelog][changelog].
52
63
53
64
## Contact
54
65
55
-
For questions and help requests, you can reach out in the [scverse discourse][scverse-discourse].
66
+
For questions and help requests, you can reach out in the [scverse Discourse][link-discourse].
56
67
If you found a bug, please use the [issue tracker][issue-tracker].
57
68
58
69
## Citation
@@ -73,7 +84,6 @@ References for individual metrics can be found in the corresponding documentatio
0 commit comments