Closed benp175 closed 3 years ago
After doing some testing fitting with the new data transformation scheme, I noticed that spin angles weren't properly being transformed. Upon further digging, I found the following code:
https://github.com/dragozzine/multimoon/blob/7571dc8f2a0f54be6f13397829ea2c6066cfc5e7/src/mm_param.py#L87-L92
By default, the primary can never have its spin angles transformed by this code. This just needs to be modified to apply to the primary.
Likewise, the transformations will have to be changed. With the specific lines shown below.
https://github.com/dragozzine/multimoon/blob/7571dc8f2a0f54be6f13397829ea2c6066cfc5e7/src/mm_param.py#L288-L305
Fixing this will significantly enhance the effectiveness of the MCMC sampling in many cases.
Looks like a commit from @dallinspencer fixes this. Closing the issue.
After doing some testing fitting with the new data transformation scheme, I noticed that spin angles weren't properly being transformed. Upon further digging, I found the following code:
https://github.com/dragozzine/multimoon/blob/7571dc8f2a0f54be6f13397829ea2c6066cfc5e7/src/mm_param.py#L87-L92
By default, the primary can never have its spin angles transformed by this code. This just needs to be modified to apply to the primary.
Likewise, the transformations will have to be changed. With the specific lines shown below.
https://github.com/dragozzine/multimoon/blob/7571dc8f2a0f54be6f13397829ea2c6066cfc5e7/src/mm_param.py#L288-L305
Fixing this will significantly enhance the effectiveness of the MCMC sampling in many cases.