Notes on existing code; reviewing notebooks and scripts

[bnord]

This discussion is just notes by bnord to review the existing code and understand it. I won't tag anyone here. Feel free to mute (if possible) or ignore this discussion.

I'm looking through all the code at this link: https://github.com/deepskies/SimSZ/blob/dev-massdist/

[bnord]

Mass Sampling Notebooks

notebook list

1. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/mass_sampling.ipynb
2. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/mass_sampling_demo.ipynb
3. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/mass_sampling_functions.ipynb
4. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/mass_sampling_functions_fin.ipynb

comments

1. These create a mass distribution drawn from known observed/empirical mass functions.
2. These will not be used for alpha release because they are not needed for training models. They can be ignored for alpha release.

[bnord]

Angular Diameter Distance Notebooks

notebook list

1. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/angular%20diameter%20distance.ipynb
2. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/angular%20diameter%20distance2.ipynb

Comments

Summary of distance2 notebook.

1. I couldn't get the first notebook to load: "invalid notebook"
2. The first part of these notebooks just tests the astropy angular diameter distance method
3. The second part applies the following steps
   4. creates a mass-redshift grid
   5. creates beam noise
   6. use angular diameter distance function to get a radius
   7. put radius into the calculation of a battaglia profile
   8. turn the battaglia profile into a y profile
   9. generate an image
   10. plot some profiles

[bnord]

Assessment

1. this notebook uses many elements that we would put into a differently designed pipeline.
2. It was good work by the students
3. we do not need to use this notebook directly.
4. We can make our pipeline in a more pythonic and code-rigorous way.

[bnord]

Noise Analysis

Notebook list

https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/noise_analysis.ipynb

Comments

Summary of notebook

1. does almost the same as the distance2.ipynb
2. except it returns "noise" from a simulator - sim2.generate_img - units unknown
3. then it plots some kind of average noise difference

[bnord]

Simulators

Notebook list

1. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/generator.ipynb
2. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/generator2.ipynb
3. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/generator_F2.ipynb
4. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/generator_h5.ipynb

Comments

Summary of Generator Notebook

1. loops over mass to create profiles and images, similar to noise analysis and angular diameter distance notebooks
2. plots annuli on top of the image

Summary of Generator2 Notebook

1. Similar to Generator Notebook, but unfinished

Summary of Generator F2 Notebook

1. unable to render

Summary of Generator h5 Notebook

1. reads an h5: masses, temperature submap, other things
2. puts into an npy
3. calculates ynoise from temp submap
4. creates sz signal from redshift and temp submap
5. calculates aperture from sz signal

I'm not sure why there's a temperature submap already in the h5

[bnord]

Simulation notebooks

Notebook list

1. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation.ipynb
2. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation_2.ipynb
3. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation_3.ipynb
4. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation_4.ipynb
5. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation_5.ipynb
6. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/Simulation_py.ipynb

Comments

Simulation

1. all of the major functions for the physics, but no mass sampling. also plots

Simulation 2

1. similar to Simulation, but plots a profile from center of image instead of corner
2. adds noise to plot
3. adds cmb to plot

Simulation 3

looks like a cleaner version of Simulation 2

Simulation 4

Same as simulation 3 but allows the mass and redshift to be more easily changed; basically makes them variables outside the functions, I think

Simulation 5

same as simulation 4 but has more 1d radial profiles

Simulation py

1. does some of the stuff from Simulation 5, but doesn't make plots
2. reads in from yaml files.

Assessment

if anything is used again, it should just be the yaml reading and all the function in Simulation 5

[bnord]

ACT DR5

Notebook list

1. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/ACTDR5_F1.ipynb
2. https://github.com/deepskies/SimSZ/blob/dev-massdist/notebooks/ACTDR5_F1_2.ipynb

Comments

I can't load these in the web browser yet. Will come back to it.