Skip to main content

Data Preparation

AnoFox Forecast provides 14 data preparation functions across 4 categories: gap filling (2 functions for missing timestamps), series filtering (4 functions to remove constant, short, gappy, or all-zero series), edge cleaning (3 functions for leading/trailing zero removal), and imputation (5 functions including forward fill, backward fill, mean fill, constant fill, and differencing). These functions operate on grouped time series data, processing all series in a single SQL call. Clean data is a prerequisite for reliable forecasts -- gaps and structural artifacts in the input directly degrade model accuracy.

Clean, transform, and prepare your time series data for forecasting.

Categories

All Functions

FunctionCategoryDescription
anofox_fcst_ts_fill_gaps_byGap FillingFill missing timestamps with NULL values
anofox_fcst_ts_fill_forward_byGap FillingExtend series to target date
anofox_fcst_ts_drop_constantFilteringRemove series with no variation
anofox_fcst_ts_drop_shortFilteringRemove series below minimum length
anofox_fcst_ts_drop_gappyFilteringRemove series exceeding max gap ratio
anofox_fcst_ts_drop_zerosFilteringRemove all-zero series
anofox_fcst_ts_drop_leading_zerosEdge CleaningRemove initial zero values
anofox_fcst_ts_drop_trailing_zerosEdge CleaningRemove trailing zero values
anofox_fcst_ts_drop_edge_zerosEdge CleaningRemove both leading and trailing zeros
anofox_fcst_ts_fill_nulls_constImputationReplace NULLs with constant value
anofox_fcst_ts_fill_nulls_forwardImputationForward fill (LOCF)
anofox_fcst_ts_fill_nulls_backwardImputationBackward fill (NOCB)
anofox_fcst_ts_fill_nulls_meanImputationFill with series mean
anofox_fcst_ts_diffImputationDifferencing transformation
Showing 14 of 14
🍪 Cookie Settings