CCL now support computing the non-Gaussian covariance. We need to interface it to TJPCov
Contacts: @carlosggarcia
Day/Time: Thursday, Oct 31 at 9-12am UK time
Main communication channel: Slack
GitHub repo: https://github.com/LSSTDESC/TJPCov/issues/106
Zoom room (if applicable):
Goals and deliverable
Implement the non-Gaussian covariance in TJPCov
Resources and skills needed
Python and CCL and a bit of halo model.
Detailed description
Apart from building the interface, we might need to do some approximations as we did with the SSC to avoid having to specify halo profiles, etc.
Add non-Gaussian covariance to TJPCov
CCL now support computing the non-Gaussian covariance. We need to interface it to TJPCov
Contacts: @carlosggarcia Day/Time: Thursday, Oct 31 at 9-12am UK time Main communication channel: Slack GitHub repo: https://github.com/LSSTDESC/TJPCov/issues/106 Zoom room (if applicable):
Goals and deliverable
Implement the non-Gaussian covariance in TJPCov
Resources and skills needed
Python and CCL and a bit of halo model.
Detailed description
Apart from building the interface, we might need to do some approximations as we did with the SSC to avoid having to specify halo profiles, etc.