Clarifying Export-to-Excel interface

[kdahlquist]

I'd like to clarify the language of the Export-to-Excel modal.

The top part should say:
Select the edge type

• Unweighted
• Weighted

Instead of not showing weighted when it's not possible, let's show it, but have it grayed out.

For the bottom part:
The current data should be the top selection, but for the rest, could they be in alphabetical order like they appear in the expression data drop-down menu?

In my mind, if I select expression data from the drop-down in the node menu, that is the option that should be selected when I chose export data to excel. For example, if I load demo #1, change the data to Barreto and then click export, Barreto data should be the one selected, not demo 1.

[kdahlquist]

The second modal then needs to be dependent on the first. If I chose Barreto data, then that is the only option for an expression data sheet that I should be given. It looks like it is adding the Dahlquist data to this.

[kdahlquist]

I would also generalize the second modal so that the use can select whether or not to export any of the worksheets:

• degradation_rates
• production_rates
• wt_log2_expression
• any other additional _log2_expression sheets
• network_weights
• optimization_parameters
• threshold_b

Note that for non-Dahlquist data, they are all wt for now, so for Barreto, for example, it would just be the one sheet "wt_log2_expression"

Note also, the "network_weights" sheet is in addition to the "network" sheet. It is a duplicate with the same name, but is necessary for GRNmap to work.

[kdahlquist]

Also, regarding the placement of the buttons.

Should the Continue button go down and to the left of the Close button?

Also, maybe the Close button should be labeled Cancel

[kdahlquist]

For the second modal, move the Export button to the left of the "Close" button, but rename "Close" to "Cancel". Also, since there are two modal windows in a series, should there be a "back" button to take you back to the previous one?

[Onariaginosa]

I started making headway on this refactor / clarification

• I moved the buttons and added the back button all into the footer. I renamed close to cancel.
• I also fixed the behavior where if you are exporting the sheet you loaded in it would concatenate all the sheets loaded during the current session. Now it only contains the files from the specific source (Hopefully)

[kdahlquist]

So far so good. Let me know when beta is refreshed and I can look at it. What does "All expression sheets" mean?

[Onariaginosa]

Progress report on UI changes

[kdahlquist]

Some notes:

• Demos 3 & 4 are are in a deprecated format. It's OK to export them "as is" but they may not necessarily work with the current GRNmap implementation.
• For "Network Sheets", there would be a total of three options:
  • "network" would contain the base network
  • "network_weights" would be an identical copy of the "network" sheet. This is meant to be the initial guesses for the optimization. Technically, the numbers could be different than 1, but in practice we just set the initial guess to 1.
  • "network_optimized_weights" will only be present if the export comes from a demo or the user uploaded a GRNmap export workbook in the first place.
• For "Expression Sheets", only Dahlquist data is going to have data for different strains. For all the other datasets, use "wt_log2_expression".
• For "Additional Sheets"
  • "degradation_rates" should come from the database.
  • "production_rates" are the initial guesses for the optimization. We use 2X the degradation rate.
  • "threshold_b" are initial guesses and can be set to "0"
  • "optimization_parameters"
    • For alpha, kk_max, MaxIter, TolFun, MaxFunEval, TolX, production_function, L_Curve, estimate_params, make_graphs, fix_P, fix_b, the default values from demo 1 or 2 can be used.
    • "expression_timepoints" needs to be figured out from what is in the "wt_log2_expression" sheet.
    • "Strain" is only going to be "wt" if non-Dahlquist data. Otherwise, needs to be figured out based on which expression sheets are ther.
    • "simulation_timepoints" are going from 0 to the last expression timepoint, in increments of 5 minutes

[kdahlquist]

One more note:

Anything that says "optimized" in the sheet name, came from a GRNmap output file, either from a demo or uploaded to GRNsight. Possibilities include:

• "network_optimized_weights"
• "wt_optimized_log2_expression" (or other strain prefixes)
• "optimized_production_rates"
• "optimized_threshold_b"
• "optimization_diagnostics"

Then there are the "sigmas" as in "wt_sigma", etc. These would only come from a demo or an uploaded GRNmap output workbook.

We should account for all of these so that we have a complete circular upload/export of identical data.

[kdahlquist]

@Onariaginosa, let me know if you have questions!

[Onariaginosa]

So, I have began working on fixing the export functionality for this feature. The UI is essentially complete, but the back end exporting functionality doesn't work as it should yet. Coincidentally, I ended up fixing the bug where a network with no regulatory connections crashes GRNsight because it was messing up the export process. Because I couldn't find an open issue it is now referenced as issue #990. As the export to excel functionality is currently wonky, I cannot submit a pr quite yet.

[Onariaginosa]

So export functionality generally works with the new interface. Now I need to focus on generating production rate, degradation rate, and optimization parameter files from our database for export purposes. The export functionality changes are located in WIP PR #985

[kdahlquist]

This looks done; some bug testing could be done as part of issue #1012