statsmodels.tsa.statespace.kalman_filter.KalmanFilter.diff_endog

KalmanFilter.diff_endog(new_endog, tolerance=1e-10)

Compute indices of revised and newly added observations

Parameters:

new_endogarray_like

New observation data to compare against the currently bound endogenous data.

tolerancefloat, optional

Tolerance used when comparing values to determine whether a change represents a revision. Default is 1e-10.

Returns:

revision_ixlist of tuple

List of (row, column) indices indicating which values in new_endog represent a revision to a previously observed (non-missing) value.

new_ixlist of tuple

List of (row, column) indices indicating which values in new_endog represent a newly observed value (i.e. one that was previously missing).