python - Curve_fit not converging means...? -

python - Curve_fit not converging means...? -

i need crossmatch list of astronomical coordinates different catalogues, , want decide maximum radius crossmatch. avoid mismatches between list , catalogues.

to this, compute separation between best match catalogue each object in list. initial list supossed position of known object, happend not detected in catalog, , coordinates may suffer little offsets.

they way computing maximum radius fitting gaussian kernel density of separation gaussian, , utilize center + 3sigmas value. method works nicely of cases, when little subsample of list has offset, have 2 gaussians instead. in these cases, specify max radius in different way.

my problem when happens, curve_fit can't fit 1 gaussian. scientific publication, need justify "no fit" in curve_fit, , in cases "different way" used. give me hand on means in mathematical terms?

there varying lengths can go justifying or fitting ansatz --- depends on details of specific case (eg: why expect gaussian work in first place? depth need/want delve why fitting procedure fails , fail etc).

if question curve_fit , failure converge, show code , input info demonstrate problem.

if question how evaluate goodness-of-fit, you're best off going library , picking book on statistics.

if way of justifying why in case gaussian not fitting ansatz, 1 way calculate moments: gaussian distribution 1st, 2nd, 3rd , higher moments related each other in precise way. if can demonstrate underlying info relation between moments different, sounds reasonable these info can't fit gaussian.

python curve-fitting