Jan, Deirdre, Soumen, Sayantani, Parthapratim, Poulomi, Saleem, Sathya, Arun and Ish
## Agenda
## Minutes
- Minute taker: Ish
- Minutes from last call: [link](https://git.ligo.org/sayantani.datta/pca_tgr/-/wikis/June%2030,%202023)
Sayantani- Ran into some troubles with the GR injections, so showing the non-GR results right now.
Combining non-GR injections- Restricted
Ish- In the plot, phi_cap_PCA_6 gives a value other than 0. Is that okay?
- Sayantani- Since most of the information is contained in the first two parameters and all the noise is dumped into the latter. So, we should not even look at these parameters.
Combining non-GR injections- Heirarchial
Arun- How close is the obtained Gaussian distribution with the injected one?
- Sayatani- The mu parameter will not have a meaning for a power-law distribution, but sigma seems to be consistent with what is expected.
Sathya- It is interesting that while all the injected deviation parameters are +ve, when you construct the PCA parameters, some of them are +ve and some are -ve. But we don't see a trend (right now). Need to dig deeper into this.
- Jan- Completely fine in principle, as PCA parameters are just linear transformations.
## Action items
1. Need to explain the positive and negative-ness of the PCA parameters.
