Skip to content

Commit

Permalink
Specify interpret package version and fix links in docs (#126)
Browse files Browse the repository at this point in the history
  • Loading branch information
navdeep-G authored Jan 15, 2021
1 parent 376c915 commit 8651e2c
Show file tree
Hide file tree
Showing 2 changed files with 5 additions and 4 deletions.
7 changes: 4 additions & 3 deletions explainers/doc/CREATING_CUSTOM_EXPLAINER_WITH_MLI_BYOR.md
Original file line number Diff line number Diff line change
Expand Up @@ -42,10 +42,11 @@ When MLI user starts interpretation, model compatible explainers (from the avail
### What is the role of recipes?
BYOR allows Data Scientists to bring their **own recipes** or leverage the existing, **open-source** recipes to explain models. In this way, the expertise of those creating and using the recipes is leveraged to focus on domain-specific functions to build customizations.
### Where can you find open-source recipes?
The curated open-source recipes currently reside in the [Responsible ML GitHub repository](https://github.com/h2oai/responsible-ml) GitHub repository:
The curated open-source recipes currently reside [here](https://github.com/h2oai/driverlessai-recipes/tree/rel-1.9.1/explainers/explainers):

* [Morris sensitivity analysis](https://github.com/h2oai/driverlessai-recipes/blob/rel-1.9.1/explainers/explainers/morris_sensitivity_explainer.py) explainer
* explainer [templates](https://github.com/h2oai/driverlessai-recipes/tree/rel-1.9.1/explainers/explainers/templates)

* [Morris sensitivity analysis](https://github.com/h2oai/h2oai/blob/dev/tests/test_mli/byor/morris_sensitivity_explainer.py) explainer
* explainer [templates](https://github.com/h2oai/h2oai/tree/dev/h2oaicore/mli/byor/recipes/mock)
## Custom Explainer
![overview](images/CREATING_CUSTOM_EXPLAINER_WITH_MLI_BYOR.overview.png)

Expand Down
2 changes: 1 addition & 1 deletion explainers/explainers/morris_sensitivity_explainer.py
Original file line number Diff line number Diff line change
Expand Up @@ -38,7 +38,7 @@ class MorrisSensitivityLeExplainer(CustomExplainer, CustomDaiExplainer):
# declaration of explanation types this explainer creates e.g. feature importance
_explanation_types = [GlobalFeatImpExplanation]
# Python package dependencies (can be installed using pip)
_modules_needed_by_name = ["interpret"]
_modules_needed_by_name = ["interpret==0.1.20"]

# explainer constructor must not have any required parameters
def __init__(self):
Expand Down

0 comments on commit 8651e2c

Please sign in to comment.