interpolated_timeseries

eolab.georastertools.processing.algo.interpolated_timeseries(dates: MaskedArray, series: MaskedArray, output_dates: array, nodata) → ndarray[source]

Interpolate a timeseries of data. Dates and series must be sorted in ascending order.

Parameters:

  • dates (numpy.ma.masked_array) – A masked array of timestamps (dates) corresponding to the input series. Should be in ascending order.

  • series (numpy.ma.masked_array) – A list of 3D masked arrays, each with shape (bands, height, width), containing the raster data for each timestamp in dates.

  • output_dates (numpy.array) – A 1D array of timestamps for which to generate the interpolated rasters.

  • nodata (float) – Value to use for pixels where input data is NaN or missing.

Returns:

A 4D numpy array of shape (time, bands, height, width), containing

the interpolated raster data for each output date. If there are no valid data points for a specific pixel, the corresponding pixel will be filled with nodata.

Return type:

numpy.ndarray

Raises:

ValueError – If series is empty, or if dates and series dimensions do not match.