astropy/halotools

Improve testing of rp_pi_tpcf_jackknife

Closed this issue · 0 comments

Currently there are no science validation tests of rp_pi_tpcf_jackknife, only unit-tests enforcing that no exceptions are raised when calling the function, checks on dimensional correctness of the returned values, etc.

Thanks to a recent science validation test of wp_jackknife provided by @yymao, who has an analytical calculation replacing halotools use of randoms, the wp_jackknife function has a reasonable level of independent science validation done for it. It should be straightforward to write a unit-test enforcing that wp_jackknife and rp_pi_tpcf_jackknife return the same covariance matrix for appropriately chosen bins.

Currently working on this with @duncandc