Adding table overview with frequency options, supersedes PR #1092 #1095
Adding table overview with frequency options, supersedes PR #1092 #1095yoselalberto wants to merge 21 commits into
Conversation
The documentation described freq as "a string indicating the frequency" but it also accepts an integer for integer-indexed time series. Closes Nixtla#986
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
below is an screenshot of the changes, they are almost identical in both Pandas and Polars: |
|
I rendered the documents with the changes locally, everything seems to be fine: |
|
I rendered the documents with the changes locally, below are screenshots, everything seems to be fine: |
…om formeo repo; adapted from https://github.com/formeo/statsforecast/tree/feat/ucm-model, original authors: Gordienko Roman (formeo@yandex.ru) and Mariana Menchero (@mmencherogarcia)
…s (values are numpy.ndarray's)
|
Hello Nixtla, thanks for your attention; after examining the failed CI/CD jobs I added the following changes: 1.- build-docs job: I specified the full quarto version used to build the docs, as required 2.- test-m3-performance job: I pulled the latest version of datasetsforecast from github, which fixes the m3 downloading; and added local functions to calculate the losses, as those imported from from utilsforecast.losses were failing because they now expect dataframes instead of numpy.arrays |
|
Hello Nixtla, good afternoon I noticed that statsforecast was updated, I will incorporate the latests changes, and update the PR so that it integrates smoothly |
|
I integrated the latest changes in main, now it should merge smoothly |
2 participants
Problem: the
freqparameter in StatsForecast is critical, yet there is no reference table; I created a frequency reference table and I added it to the Getting Started guides for both Pandas and Polars; this PR hopes to reduce friction while using StatsForecast.In addition, I am thinking about creating a post with a getting started guide with examples for various frequency options, and a section with troubleshooting guide for common frequency errors; what do you think about that @nasaul @MMenchero? are you willing to add a new post?
Supersedes PR #1092 by @W057; closes issues #996, and it is related to #986.
Notes: I updated the PR, integrating the latest changes in main; in addition, I updated datasetsforecasts install, now it is resolved from PyPI instead of git