Prescriptor select #10
Conversation
There was a problem hiding this comment.
Updated readme to reflect new project structure
| @@ -32,7 +31,7 @@ | |||
| app.title = "Climate Change Decision Making" | |||
|
|
|||
| context_component.register_callbacks(app) | |||
| parallel_component.register_callbacks(app) | |||
| filter_component.register_callbacks(app) | |||
There was a problem hiding this comment.
Replaced parallel coordinates component with new filter component
There was a problem hiding this comment.
Created some classes we can re-use between components
| Starts by plotting "other" so that it's the bottom of the z axis. | ||
| Then plots selected candidates in color. | ||
| Finally plots the baseline on top. | ||
| """ |
There was a problem hiding this comment.
Moved parallel coordinates code into filter.py
| html.Div( | ||
| dbc.Accordion( | ||
| dbc.AccordionItem(dcc.Graph(id="parcoords-figure"), title="View Parallel Coordinates"), | ||
| start_collapsed=True, |
There was a problem hiding this comment.
Hide parallel coordinates in accordion dropdown
| ), | ||
| dbc.Col( | ||
| dbc.Button( | ||
| "Explore & Fine-Tune Policy in En-ROADS", |
There was a problem hiding this comment.
Big fancy button to open En-ROADS to draw user in
|
|
||
| if "baseline" in cand_idxs: | ||
| baseline_outcomes_df = outcomes_dfs[-1] | ||
| baseline_outcomes_df["year"] = list(range(1990, 2101)) |
There was a problem hiding this comment.
We no longer can toggle baseline nor other
| @@ -13,6 +13,27 @@ | |||
| from evolution.outcomes.outcome_manager import OutcomeManager | |||
|
|
|||
|
|
|||
| def filter_metrics_json(metrics_json: dict[str, list], | |||
| metric_ranges: list[tuple[float, float]], | |||
| normalize=False) -> pd.DataFrame: | |||
There was a problem hiding this comment.
Moved this function into utils so it can be used in filtering too.
| outcomes_dfs: list[pd.DataFrame]) -> pd.DataFrame: | ||
| """ | ||
| Takes parallel lists of context_actions_dicts and outcomes_dfs and processes them into a metrics dict. | ||
| All of these metrics dicts are then concatenated into a single DataFrame. |
There was a problem hiding this comment.
This is nice for filtering
Created a nice, interactive way to filter out prescriptors to acquire desired behavior.
Refactored page to be nicer formatted via. bootstrap.