Adding example for testing proportions #581
Conversation
|
Thanks for contributing this nice example, @freddyaboulton ! I'll review this in detail tomorrow. |
| impact_on_probability = response_with_calls - baseline_response | ||
|
|
||
| print(f"There is a {interval_size * 100}% probability that calling customers " | ||
| "increases the chance they'll make a purchase by " |
There was a problem hiding this comment.
This looks good to me since we constructed the dataset, but users extrapolating it to more complex examples with multiple predictors might be problematic. @fehiepsi, @martinjankowiak: WDYT?
There was a problem hiding this comment.
i think this is fine for this self-contained example 👍
| impact_on_probability = response_with_calls - baseline_response | ||
|
|
||
| print(f"There is a {interval_size * 100}% probability that calling customers " | ||
| "increases the chance they'll make a purchase by " |
There was a problem hiding this comment.
i think this is fine for this self-contained example 👍
| :param design_matrix: Covariates. All categorical variables have been one-hot | ||
| encoded. | ||
| :param outcome: Binary response variable. In this case, whether or not the | ||
| customer made a purchase. | ||
| """ | ||
|
|
||
| # Use multivariate normal prior in case we want to add more covariates later. | ||
|
|
||
| beta = numpyro.sample('coefficients', dist.MultivariateNormal(loc=0., | ||
| covariance_matrix=np.eye(design_matrix.shape[1]))) |
There was a problem hiding this comment.
nit: this can be faster if you replace covariance_matrix by scale_tril. It is still fine if you want to leave it as-is.
examples/proportion_test.py
Outdated
|
|
||
| ) | ||
|
|
||
| print("There is a 95% probability the effect of gender on the log odds of conversion " |
There was a problem hiding this comment.
This is still not addressed yet.
|
@fehiepsi @martinjankowiak @neerajprad I think I've addressed your comments! Please let me know if you need me to change anything else. |
examples/proportion_test.py
Outdated
|
|
||
| ) | ||
|
|
||
| print("There is a 95% probability the effect of gender on the log odds of conversion " |
There was a problem hiding this comment.
This is still not addressed yet.
|
@fehiepsi Thank you for the help! Should be all good now. |
In reference to issue #189, I added an example for testing if two proportions are equal using logistic regression. This example uses a self-contained, simulated dataset.