Repository Issue Activity (beta)

#1235 AutoARIMA: drop collinear exogenous regressors instead of fitting them

#199 Multiple seasonal arima

#1230 GARCH/ARCH prediction intervals are computed as `mean ± z·sigma2` (variance) instead of `mean ± z·sqrt(sigma2)`

#1221 [Core] ConformalIntervals accepts h=0 and later raises ZeroDivisionError in _conformity_scores

#485 [Models: Automatic BoxCox estimation inside AutoARIMA]

#409 Enable blamda for arima

#1214 Seasonality and trend extraction in the presence of level shifts

#1198 spark.createDataFrame(series) error

#1202 [Models] ConformalSeasonalPool: documented n_samples sufficiency rule covers only the lower bound and its example is off by one; no validation or warning when the pool is too small

#753 remove numba dependency

#1191 [Enhancement] Support empirical bootstrap prediction intervals in `forecast_arima` and `forecast_ets`

#1182 Add support for pandas>=3.0

#1190 uv add statsforecast Error with python 3.12

#1164 [FEAT] Add conformal_error method for prediction intervals

#1167 ARIMA(2,0,0) returns unexpected coefficient fits and prediction

#1187 Python 3.14 wheels

#1121 feat: Consider adding `py.typed` file marker for type-checking support

#1077 [Python 3.14 errors with scipy installation]

#1119 Conformal intervals become extreme/null after inverse log transform (expm1) — what's the recommended approach?

#1172 Scipy Upgrade

#1154 [StatsForecast:MSTL] - index out of bounds error for period array with length greater than 6

#1135 theta: NaN forecast intervals when len(y) == 4

#870 [AutoARIMA] when using AutoARIMA on a constant time series the forecast fitted values are zero no matter the constant.

#482 negative values in confidence levels for prediction intervals should not be allowed

#1140 SeasonalNaive: forecasts misalign with the seasonal period when len(y) < season_length