Inference
Inference(self, alpha, p, tests=1)Base class for inference.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| alpha | float | Probability of Type I error \(\alpha\). | required |
| p | int | Number of parameters \(p\). | required |
| tests | int | Number of hypothesis tests. | 1 |
Attributes
| Name | Description |
|---|---|
| alpha | Probability of Type I error \(\alpha\). |
| conf_int | Confidence intervals for each parameter (\(p \times 2\) matrix). |
| estimate | Estimates for each parameter. |
| n | Number of observations \(n\). |
| names | Names for each parameter. |
| p | Number of parameters \(p\). |
| p_value | P-values for each hypothesis test. |
Methods
| Name | Description |
|---|---|
| batch | For each sample unit in the batch |
| infer | Calculate confidence interval and p-value, then |
| update | Update statistics with new data. |
batch
Inference.batch(xs, **kwargs)For each sample unit in the batch
- call
update, - call
infer, and - append the
Inferenceobject to a list.
Returns
| Name | Type | Description |
|---|---|---|
| List[Inference] |
infer
Inference.infer(**kwargs)Calculate confidence interval and p-value, then
- keep the maximum lower bound and minimum upper bound for the confidence interval; and
- keep the minimum p-value.
update
Inference.update(x, **kwargs)Update statistics with new data.