Skip to content

Add NumPy, pandas, and CSV forecasting support for 0.2.0 - #8

Open
Benjamin-Walker wants to merge 1 commit into
mainfrom
upgrade-0.2.0-input-formats
Open

Benjamin-Walker wants to merge 1 commit into
mainfrom
upgrade-0.2.0-input-formats

Conversation

@Benjamin-Walker

Copy link
Copy Markdown
Contributor

Allow model.forecast to have NumPy, pandas and CSV inputs. NumPy inputs return arrays, pandas/CSV inputs return DataFrames, and Torch outputs stay on the input device. Dated inputs supply calendar features and future forecast dates.

Bumps the package to 0.2.0 and updates the usage documentation with an example CSV.

Verified with 394 tests and 3 package checks; two release-weight tests skipped. Lint, formatting, and build checks pass.

Preserve input tensor devices and return DataFrames for pandas and CSV inputs. Infer calendar metadata and future date indexes from dated histories.

Bump the package to 0.2.0, add format and date regression tests and an example CSV, and update the usage documentation.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant