phys201 / FluorescenceSpectroscopy

Bayesian inference modeling for fluorescence spectroscopy

GNU General Public License v3.0

1 stars 1 forks source link

sampler tweaks #5

Closed vnmanoharan closed 2 years ago

vnmanoharan commented 2 years ago

For NUTS sampling, you don't want to start at the MAP because the gradients are zero there. So I'd recommend removing the start=pm.find_MAP() argument in your pm.sample() statement. Also it would be good to see some comparison between the inferred and true values for your simulated data.

bdelwood commented 2 years ago

8 was only a partial fix, still have

it would be good to see some comparison between the inferred and true values for your simulated data.

to do

bdelwood commented 2 years ago

Fixed in b62f86f00a85700bcdc590e4896779961dcad3f3 by final plot comparing inferred from real parameter values.