Non poissonian tests #205
Non poissonian tests #205
Conversation
removed unused imports
| if random_numbers is None: | ||
| random_numbers = numpy.random.rand(num_events) | ||
| else: | ||
| # TODO: ensure that random numbers are all between 0 and 1. | ||
| pass | ||
|
|
||
| # reset simulation array to zero, but don't reallocate | ||
| sim_fore.fill(0) |
There was a problem hiding this comment.
@bayonato89 @khawajasim can we remove this part of the function? the vector random_numbers is not used for anything it seems. the simulation below allocates a new array called random_num. i think we still need to keep the sim_fore.fill(0) should be included.
csep/core/binomial_evaluations.py
Outdated
| if (idx + 1) % 100 == 0: | ||
| print(f'... {idx + 1} catalogs simulated.') | ||
|
|
||
| target_idx = numpy.nonzero(catalog.ravel()) |
There was a problem hiding this comment.
this line also doesn't work because the variable catalog is not defined anywhere here. should this be observed_data?
|
thanks for this contribution. looks like we're getting really close. i left some comments on this pull request. we still need a test case here to verify that everything is working as we expect. @khawajasim @bayonato89 can you get some test cases for this module? we need to make sure that |
|
I think I can take care of unit tests for _binary_likelihood_test, _simulate_catalogs in the coming days. |
| random_num = numpy.random.uniform(0,1) | ||
| loc = numpy.searchsorted(sampling_weights, random_num) | ||
| if sim_fore[loc] == 0: | ||
| numpy.add.at(sim_fore, loc, 1) |
pyCSEP Pull Request Checklist
Please check out the contributing guidelines for some tips
on making pull requests to pyCSEP.
Fixes issue #(please fill in or delete if not needed).
Type of change:
Please delete options that are not relevant.
Checklist: