Add support for EIA forecasts endpoint #247
Conversation
dufia
changed the title
vortexasdk/api/eia_forecast.py
Outdated
|
|
||
|
|
||
| @dataclass(frozen=True) | ||
| class EIAForecast(IDName, FromDictMixin): |
There was a problem hiding this comment.
This EIAForecast class currently inherits from IDName - which should only be the case if each EIAForecast record returned by the API has both an id attribute and a name attribute (https://github.com/VorTECHsa/python-sdk/blob/master/vortexasdk/api/shared_types.py#L41). @dufia This is likely the cause of the failing tests
There was a problem hiding this comment.
We'll also want to update the about-endpoints page that lists the available endpoints.
just notices the cargo time series endpoint is missing - while we're here can we add that?
http://localhost:8000/endpoints/about-endpoints/
https://github.com/VorTECHsa/python-sdk/blob/master/docs/endpoints/about-endpoints.md
| def search( | ||
| self, | ||
| preset: str = None, | ||
| filter_time_min: str = False, |
There was a problem hiding this comment.
The type here should be datetime, rather than string. We'll also want sensible default datetimes rather than the bool False
| - Vortexa's data science based forecast of the EIA number to be published on the week | ||
| - Actual EIA import/export numbers as published by the EIA Weekly Supply Estimates report | ||
| - EIA stocks (kbl) | ||
| - Cover (days of Supply for the whole of the US, as published by the EIA Weekly Supply Estimates report) | ||
| - Refinery runs (refiner “Percent Operable Utilization” as published by the EIA Weekly Supply Estimates report) |
There was a problem hiding this comment.
Nice - should we make this even clearer by specifying the column name first, then the accompanying documentation? something like:
runs: Refinery runs (refiner...
this means it'll be easier for clients to lookup the functionality associated with a col name
| Represent EIA forecasts as a `pd.DataFrame`. | ||
|
|
||
| # Arguments | ||
| columns: The EIA forecasts features we want in the dataframe. Enter `columns='all'` to include all features. |
There was a problem hiding this comment.
| columns: The EIA forecasts features we want in the dataframe. Enter `columns='all'` to include all features. | |
| columns: The EIA forecasts columns we want in the dataframe. Enter `columns='all'` to include all columns. |
There was a problem hiding this comment.
In the world of datascience, the term features normally refers to the input features of a model. In this sense of the word, we're better off using 'column' rather than features, as these aren't input features.
| filter_time_max: str = False, | ||
| ) -> EIAForecastResult: | ||
| """ | ||
| Find EIA forecasts for a given date preset and date range. |
There was a problem hiding this comment.
Do we mean date preset here? It's more of a geography & product preset right?
There was a problem hiding this comment.
Fixed, it should've been: "Find EIA forecasts for a given preset and date range."
| | 2020-01-31T00:00:00.000Z | 454.96048964485 | 323 | 9541 | 26.5 | 65.9 | | ||
| | 2020-01-24T00:00:00.000Z | 545.453497230504 | 579 | 10461 | 25.9 | 61.5 | | ||
| | 2020-01-17T00:00:00.000Z | 510.289752707662 | 549 | 10325 | 25.2 | 64.7 | | ||
| | 2020-01-10T00:00:00.000Z | 469.841470826967 | | | | | |
There was a problem hiding this comment.
Should we give an accompanying note as to why some of the values are null? Clients will definitely ask this Q - we may as well answer it here once.
| Find EIA forecasts for a given date preset and date range. | ||
|
|
||
| # Arguments | ||
| preset: the EIA forecasts preset to be returned. Preset can be: 'padd1-gasoline-imports', |
There was a problem hiding this comment.
Should we explain what a preset is?
Adds the EIA forecast endpoint.