Add uncertainty components to the gui

[adhamrait]

Add components to change the uncertainty in alpha and beta in the gui:

[adhamrait]

Added the two separate variables and a toggle to display one or the other (currently gives an error if trying to display the inferred calibration values)

[EZoni]

Can you remind me/us again on why we wrap everything within a with html.Div():? What does this accomplish specifically? Just trying to understand if it's necessary, optional, if it should be used elsewhere/everywhere where we set up other graphical components, etc. I'm sure I haven't used it consistently, or at all, so far, so I want to make sure I understand why it's here and what it does.

[EZoni]

I can also search the answer online, I'm asking just in case you have the answer ready off the top of your head.

[adhamrait]

I wrapped this in a div so that the alpha and beta sections are separate and they're flexboxes, but looking at it now I can just use a VRow to be more consistent with other parts of the code

[adhamrait]

These changes must happen in parallel:

https://github.com/BLAST-AI-ML/synapse-bella-staging-injector/pull/4/changes#diff-d8d0422389f03d783e32e627250fe29834bd09c6361640d1ff00661dd6820034
https://github.com/BLAST-AI-ML/synapse-bella-acave/pull/3/changes#diff-d8d0422389f03d783e32e627250fe29834bd09c6361640d1ff00661dd6820034
https://github.com/BLAST-AI-ML/synapse-bella-ip2/pull/3/changes#diff-d8d0422389f03d783e32e627250fe29834bd09c6361640d1ff00661dd6820034
https://github.com/BLAST-AI-ML/synapse-bella-qed-ip2/pull/3/changes#diff-d8d0422389f03d783e32e627250fe29834bd09c6361640d1ff00661dd6820034

[EZoni]

Do you think we can find a way to write the formulas with formatting that looks more consistent, while keeping Ruff happy?

For sim to exp we have:

• Lines 19-21:

df_sim[exp_name] = (
    df_sim[sim_name] / value["alpha_guess"] + value["beta_guess"]
)

• Lines 27-30:

df_sim[exp_name] = (
    df_sim[sim_name] / value["alpha_inferred"]
    + value["beta_inferred"]
)

I think these two are okay, probably Ruff complains about having the second one on one line (does it?).

For exp to sim we have:

• Lines 61-63:

sim_dict[sim_name] = (exp_dict[exp_name] - value["beta_guess"]) * value[
    "alpha_guess"
]

• Lines 69-71:

sim_dict[sim_name] = (
    exp_dict[exp_name] - value["beta_inferred"]
) * value["alpha_inferred"]

I think we could fix these two. Maybe something like

sim_dict[sim_name] = (
    value["alpha_guess"]
    * (exp_dict[exp_name] - value["beta_guess"])
)

and

sim_dict[sim_name] = (
    value["alpha_inferred"]
    * (exp_dict[exp_name] - value["beta_inferred"])
)

[adhamrait]

I split the conversion into its own function, how does that look?

[EZoni]

I like the idea of using functions, but I wonder if they should be declared and defined in calibration_manager.py.

They would then be used directly here, and possibly reused within the same module by convert_sim_to_exp in place of

synapse/dashboard/calibration_manager.py

Line 18 in ceea73f

df_sim[exp_name] = df_sim[sim_name] / value["alpha"] + value["beta"]

and by convert_exp_to_sim in place of

synapse/dashboard/calibration_manager.py

Line 41 in ceea73f

sim_val = (exp_dict[exp_name] - value["beta"]) * value["alpha"]