You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
After playing around with ipywidgets, I have some thoughts on making them work with dfkernel.
Widgets should be shown with labels like other outputs (need to redo the widget-specific output area stuff to show the tag), not sure if having them be execute_results or display_datas is a big deal
A widget's value should be referenced just using the the name of that widget, x not x.value as is currently done. Any needs for things like links, jslinks, etc. can be done with some type of widget() call or something.
Every widget should have an observer that triggers execution if necessary. Users should also be able to use an interact_manual analog so that execution is only triggered when a button is clicked
Downstream use of a widget's value is then as simple as y = x + 4
Apparently we do something weird with ipywidgets now, not sure if we always did but a h = j.value +3 seems to cause issues because we appear to be trying to look inside the IntSlider object?
I can't confirm this is still an issue, on the most recent version this does not appear to still be occurring will leave this open in the event someone else is able to confirm.
From @dakoop on June 28, 2018 20:8
After playing around with ipywidgets, I have some thoughts on making them work with dfkernel.
execute_result
s ordisplay_data
s is a big dealx
notx.value
as is currently done. Any needs for things like links, jslinks, etc. can be done with some type of widget() call or something.interact_manual
analog so that execution is only triggered when a button is clickedy = x + 4
Copied from original issue: colinjbrown/dfkernel#42
The text was updated successfully, but these errors were encountered: