Note

THIS PROJECT HAS BEEN MERGED INTO SUEWS AND ARCHIVED.

supy.util.fill_gap_all

supy.util.fill_gap_all(ser_to_fill: Series, freq='1D', limit_fill=1, thresh_ratio=0.8) → Series[source]

Fill all gaps in a time series using data from neighbouring divisions of ‘freq’

Parameters:

  • ser_to_fill (pd.Series) – Time series to gap-fill

  • freq (str, optional) – Frequency to identify gapped divisions, by default ‘1D’

  • limit_fill (int, optional) – Maximum number of consecutive NaNs to fill. Any number less than one means no pre-gap-filling interpolation will be done.

Returns:

  • ser_test_filled (pd.Series) – Gap-filled time series.

  • Patterns

  • ——–

  • 010 (missing data in division between others with no missing data)

  • 01 (missing data in division after one with no missing data)

  • 10 (division with missing data before one with no missing data)