Skip to content

Cross validation mods clean#329

Open
elmartinj wants to merge 11 commits intoTimeCopilot:mainfrom
elmartinj:cross_validation_mods_clean
Open

Cross validation mods clean#329
elmartinj wants to merge 11 commits intoTimeCopilot:mainfrom
elmartinj:cross_validation_mods_clean

Conversation

@elmartinj
Copy link
Contributor

Adding cleaned up version of PR

elmartinj and others added 11 commits March 13, 2026 11:25
This tutorial contains an example with multiple indexes and subsequent tampering to the data in order to show resiliency and a real life use case of TC applied on cryptocurrency prices up to 2021.
Fix to the issue raised on empty dataframe, resulting from an inner merge where the existing dataframe that accumulated results and the newer one had diferent indices. A subsequent issue must be raised to either:
1 report a single model failure (on index matching)
2 fix the moirai discrepancy (only model that showed this issue)
…okenization run, first trial version. felt cute, might delete later
…cripts in order to refine the possibilites available for cross validation. This will allow user through the analyze function to determine the amount of folds desired within a process. This should be flexible because folds obv imply more comp/time costs
…ms of cross validation to be available to user. new argument to function analyze will allow user or api to determine cross validation folds as follows: 1 fold - quick version, 3 fold - std version, 5 fold - robust version. in addition to the calculations of the median of these folds in order to decide which is the best modeling approach, std deviation is provided to the llm; thus allowing for more clarity regarding the stability of all fold adjustments. final addition is regarding the pretify function, in order to display these results if requested.
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