fisher_matrix¶
- rubin_sim.maf.maf_contrib.fisher_matrix(t, t0, te, u0, fs, fb, snr, filters='ugriz', filter_name='i')¶
The Fisher (information) matrix relying on first derivatives wrt t0,te,u0,fs,fb relying on with cse optimized coefficents. This implementation for the non-linear magnification is subject to the respective underlying assumptions, i.e. relatively small uncertainties. NB the Fisher matrix is using first derivatives, alternatively one could consider second derivatives the resulting covariance matrix as inverse of the Fisher matrix is a Cramer Rao lower bound of the respective uncertainty. The function returns the full information matrix for all t.
via Markus Hundertmark markus.hundertmark@uni-heidelberg.de
- Parameters:
- tfloat
The time of observation (days usually as JD or HJD)
- u0float
The impact parameter (0 means high magnification)
- tefloat
Einstein crossing time (days)
- t0float
Time of peak (days usually as JD or HJD)
- fsfloat
Source flux (counts/s but here fs = 10.**(-0.4*mag_source) in the respective passband)
- fbfloat
Blend flux (counts/s but here fs = 10.**(-0.4*mag_blend) in the respective passband)
- snrfloat
Signal to noise ratio in order to infer the flux uncertainty