Repository navigation

feat(ai): add data classes for time series forecasting and imputation - #2873

Merged

miladh merged 18 commits into

masterfrom

ai-ts-forecasting/01-data-classes

Nov 9, 2026

Merged

feat(ai): add data classes for time series forecasting and imputation#2873
miladh merged 18 commits into
masterfrom
ai-ts-forecasting/01-data-classes

Commits on Nov 2, 2026
- feat(ai): add data classes for time series forecasting and imputation
  
  miladh
  and
  claude
  committed
  12e6a69
- fix(ai): return an empty frame from to_pandas on an empty result list
  
  miladh
  and
  claude
  committed
  22810f5
- fix(ai): use the pandas helper for the empty index in to_pandas
  
  miladh
  and
  claude
  committed
  8772cc0
Commits on Nov 4, 2026
- fix(ai): read forecast and impute results from the items array
  
  miladh
  and
  claude
  committed
  a3ec603
- refactor(ai): rename the input classes to ForecastInputDatapoint and ForecastInputTimeSeries
  
  miladh
  and
  claude
  committed
  f71110f
- refactor(ai): take ForecastInputDatapoint only as datapoints, and add a load method
  
  miladh
  and
  claude
  committed
  d04aa67
- docs(ai): explain how value and missing affect what the model sees
  
  miladh
  and
  claude
  committed
  b675bee
- docs(ai): explain how to use cohort
  
  miladh
  and
  claude
  committed
  5856c41
- refactor(ai): parse every quantile key as a level
  
  miladh
  and
  claude
  committed
  6216571
- refactor(ai): read the response metadata as the API returns it
  
  miladh
  and
  claude
  committed
  2d5e38a
- refactor(ai): look up forecast and impute results by label only
  
  miladh
  and
  claude
  committed
  6a5bbcd
- test(ai): skip the quantile classes in the unknown-key loading test
  
  miladh
  and
  claude
  committed
  cb8eba8
Commits on Nov 6, 2026
- feat(ai): include only the median in to_pandas by default
  
  miladh
  and
  claude
  committed
  9ab0e18
- Apply batched suggestions from code review
  
  miladh
  and
  haakonvt
  authored
  cd8d112
- refactor(ai): use match/case to load a ForecastInputDatapoint
  
  miladh
  and
  claude
  committed
  a66cef7
- refactor(ai): load results with _load_raw_api_response
  
  miladh
  and
  claude
  committed
  19394d0
- refactor(ai): replace quantile_levels with include_all_quantiles in to_pandas
  
  miladh
  and
  claude
  committed
  4383741
Commits on Nov 9, 2026
- refactor(ai): keep the response metadata as an object on the result lists
  
  miladh
  and
  claude
  committed
  d38ce4c

Commits

Commits on Nov 2, 2026

Commits on Nov 4, 2026

Commits on Nov 6, 2026

Commits on Nov 9, 2026