rglab / opencytovisualization Goto Github PK

to the setup tab panel

for files such as the NetCDF file, but in the future potentially also for the generated plots...

"So, it looks like the digest package might be a good option for caching.
http://cran.r-project.org/web/packages/digest/index.html

An example, would be to create a hash digests (in md5) of a list of FCS filename. If the FCS filename are modified (some added, deleted, appended, etc) we could check that against the hash object and recreate the netcdf file. I think this would be done automatically when one uses any of our module. That way we only do it once."

There's issues with the plots and the study variables provided to the user for generating plots.
Specifically, with a reproducible example:
HVTN080, selecting visitno, Stim, and PTID.
When I plot IFNg vs IL2, and select all PTIDS, ENV and negctrl1 stimulations, and all visits (2,and 5) the plots generated only show data for visit 5, presumably because data for visit 2 is not available.
I think visit 2 should be filtered out of the plot.. it doesn't make sense to show it when that level in the visit factor is empty.

I think what I'm getting at above is that selecting certain levels of study variables in the visualization should automatically filter the available levels in the other study variables. This may already be an open issue elsewhere, so feel free to merge it with that if it's the case.

with critical errors in red, warnings in yellow, and messages in green

Do not forget to modify the x-axis click listener to not clear out the y-value after that.

Add functionality to be able to swap the values of x and y axes.

first of the generated graphics, then of the input parameters combination...

for error messages

... at once on a dashboard of a project

https://www.labkey.org/issues/home/Developer/issues/details.view?issueId=16923

avaialble since Labkey revision 25411

like when the user hovers over the separator between columns or may be over the gridPanel header -> show up message about columns width resizing

Originally by @gfinak
Next, for the stimulations, after I have selected ENV and negctrl and made a plots, the combo box doesn't allow me select other stimulations.. ie.. it filters them out and I can not select more stimulations.

So this is at the moment "as expected", since this is the way things are implemented currently:

Consider the 'grand' table:

1 2

A C
A D
B C
B D

The corresponding combos would be:
1 with choices A and B
and
2 with choices C and D

If you were to select A in combo 1, the grand table would filter down to:

1 2

A C
A D

The available choices for combo 2 still would be generated/updated as C and D

At this point you can still deselect A in combo 1 and/or select B in combo 1 and change the state of the grand table, but once you are done interacting with combo 1 and start making selections in combo 2, the grand table is now in a filtered state that resulted from interaction with combo 1 (suppose that interaction was just checking A), now if we select C in combo 2, the table would get filtered further down to:

1 2

A C

The available choices for the combo 1, as can be seen, would be generated/updated as just A (not the original A and B), if we were to deselect the C in combo 2 we just picked, the table would go back to that last filtered state

1 2

A C
A D

and the available choices for the combo 1, as can be seen, would be generated/updated as still just A (not the original A and B), the choices made in the combo from the previous filtering level

Hope this makes sense.

This is the way for it to behave as originally requested: "present in the combos only the choices available at the current filtering state".

If you wanted to have the original A and B (4 choices for your actual use case above), then you would need to click "Clear All" to reset to the fresh state, otherwise, there is no mechanism to undo the filtering steps that were done in a combo, which is not the last one interacted with.

Raphael did suggest to not filter the available choices for these study variables combos, but to use check Boxes on them (or may be there is a way to grey out the choices not available) - we will have to think and brainstorm about this, if it is not behaving appropriately...

from a csv file, for example, for ITN data on blackrhino: /loc/no-backup/mike/ITN029ST/FCS_File_mapping.csv

Use LABKEY.Query.getSchemas(config) to check if the table is present (some additional meta data has been uploaded).

to when the analysis is first selected and not when the plotting is done, that way there should not be name collisions (gating set named 'G')

http://fancyapps.com/fancybox/#examples