Preliminary results

[dlanzieri]

This issue is to track and discuss the preliminary results of our Fisher analysis. In this notebook I started to compute the Fisher matrix and the Fisher contours by using different combinations of Wavelet scales. This is an example of how a filter map looks like for the chosen scales: Following, the Fisher constraints from the multiscale peak counts compared to constraints from the multiscale power spectrum: . N.B the Red is the power spectrum and the purple is the peak counts. I also had a look at the contribution of the peak counts for a single Starlet scale. The following results correspond to the peak counts computed on a map filtered with a single Starlet scale (images-like the ones shown above) compared to multiscale power spectrum.

1. 

2. 

3. N.B the Red is the power spectrum and the blue is the peak counts.

@EiffL Can you have a look at these results?

[EiffL]

Ah, nice :-) Two comments:

• Can you make it so that all plots are on the same scale, given by the size of the PS contours?

• Can you use ChainConsumer to make these Fisher forecasts plots? You can find an example in this notebook: https://colab.research.google.com/drive/1K8cB1h3ge3kTVut81Xnkw2kNiKFIn8HI?usp=sharing

  
  from chainconsumer import ChainConsumer
  c = ChainConsumer()
  c.add_covariance(fid_params , jnp.linalg.inv(F), 
               parameters=["$\Omega_c$", "$\sigma_8$"], name="Fisher")

fig = c.plotter.plot(figsize="column", truth=fid_params)

And one question:
- What are your questions on these plots? We designed the experiment to access exactly the same scales, what are you unsure about?