Appendix H — chapter3.analysis.model
chapter3.analysis.model
Defines a model to manage the reports generated by the machine and deep learning models.
Classes
| Name | Description |
|---|---|
| ActivityMetric | Enumeration for each metric of interest from the model reports for individual activities. Values: PRECISION, RECALL, F1, SUPPORT. |
| Filter | Class to represent a filter that can be applied to a dataframe of reports |
| Model | Enumeration to represent the ML and DL models. Values: MLP, CNN, LSTM, CNN_LSTM. |
| ModelMetric | Enumeration for each metric of interest from the model reports. Values: ACCURACY, TRAINING_TIME. |
| Source | Enumeration to represent the data sources. Values: SP, SW, FUSED. |
| TargetFilter | Enumeration for each attribute of interest from the model reports. Values: MODEL, SEATED, STANDING_UP, WALKING, TURNING, SITTING_DOWN. |
ActivityMetric
chapter3.analysis.model.ActivityMetric()
Enumeration for each metric of interest from the model reports for individual activities. Values: PRECISION, RECALL, F1, SUPPORT.
Filter
chapter3.analysis.model.Filter(self, model, source, target, metric)
Class to represent a filter that can be applied to a dataframe of reports
Attributes
| Name | Type | Description |
|---|---|---|
| model | str | Model to look for. One of libs.chapter3.model.Model |
| source | str | Source to look for. One of libs.chapter3.model.Source |
| target | str | Target to look for. One of libs.chapter3.model.TargetFilter. |
| metric | str | Metric to look for. One of libs.chapter3.model.ModelMetric or libs.chapter3.model.ActivityMetric |
Methods
| Name | Description |
|---|---|
| apply | Function to apply the filter to a specified dataframe. |
apply
chapter3.analysis.model.Filter.apply(df)
Function to apply the filter to a specified dataframe.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
df |
pandas.DataFrame |
Model reports. | required |
Returns
| Type | Description |
|---|---|
pandas.DataFrame |
Filtered model reports. |
Model
chapter3.analysis.model.Model()
Enumeration to represent the ML and DL models. Values: MLP, CNN, LSTM, CNN_LSTM.
ModelMetric
chapter3.analysis.model.ModelMetric()
Enumeration for each metric of interest from the model reports. Values: ACCURACY, TRAINING_TIME.
Source
chapter3.analysis.model.Source()
Enumeration to represent the data sources. Values: SP, SW, FUSED.
TargetFilter
chapter3.analysis.model.TargetFilter()
Enumeration for each attribute of interest from the model reports. Values: MODEL, SEATED, STANDING_UP, WALKING, TURNING, SITTING_DOWN.
Functions
| Name | Description |
|---|---|
| obtain_best_items | Determines the best performant item (model or data source) from a DataFrame of statistical tests comparing groups. |
obtain_best_items
chapter3.analysis.model.obtain_best_items(test_results, focus_on, groups)
Determines the best performant item (model or data source) from a DataFrame of statistical tests comparing groups.
Parameters
| Name | Type | Description | Default |
|---|---|---|---|
test_results |
pandas.DataFrame |
DataFrame with statistical tests. Generated with libs.chapter3.statistical_tests.statistical_comparison. |
required |
focus_on |
list[str] | Items being compared. List items are one of libs.chapter3.model.Source or libs.chapter3.model.Model. |
required |
groups |
list[str] | Each group where focus_on items are compared. List items are one of libs.chapter3.model.Source or libs.chapter3.model.Model. |
required |
Returns
| Type | Description |
|---|---|
| dict | Dict containing the best performant item. |