Select Points -- Highlight Gated -- Show Image

cellprofiler-analyst

#1

Hi!

after I select a region in a scatterplot, the “Highlight Gated” command does not(!) only select the points within the black rectangle, but all points turn blue.

However, the text output of the program seems correct to me:

**
Select DISTINCT Per_Object.nucleus_Location_Center_X,Per_Object.nucleus_Location
_Center_Y,Per_Object.ImageNumber FROM Per_Object WHERE (Per_Object.ImageNumber =
1 and ( 353.89412286989796 < Per_Object.nucleus_Location_Center_X ) and ( 960.0
69204617347 > Per_Object.nucleus_Location_Center_X ) and ( 212.52116148617043 <
Per_Object.nucleus_Location_Center_Y ) and ( 873.425665455588 > Per_Object.nucl
eus_Location_Center_Y ) )

Select DISTINCT Per_Object.nucleus_Location_Center_X,Per_Object.nucleus_Location
_Center_Y FROM Per_Object WHERE (Per_Object.ImageNumber = 1 and Per_Object.Image
Number in (1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1
,1,1,1,1,1,1))**

in addition, when i then click “Show Random Image from selected”, nothing happens and the Program outputs:

This method may not be correct
Exception in thread “AWT-EventQueue-0” java.lang.ClassCastException: java.lang.S
tring cannot be cast to jCPAnalyst.Visualize.Model.TableColumn

However, the connection to the database seems to work find as I can display the measured parameters without problems.

I was wondering whether it is an issue that I use WindowsXP 64Bit?!
Could that be an issue?!

Would be great if you could help!!!

Best wishes,
Christian


Scientist at the Cell Screening Facility at the EMBL Heidelberg


#2

…just checked: the same happens also with WindowsXP 32bit


#3

Hi Christian,

I thank you for your feedback on CPA in this forum, as it’s important we know what features are (1) being used, and (2) not working the way people expect them.

It looks like you are trying to plot per-object data in scatterplot. To be honest, I didn’t even think the app would let you attempt this, but I see now the results you are getting when doing it myself. Unfortunately, this feature is one that wasn’t properly implemented, and was never fixed because we decided that the scatterplot be used only for per-image data. You might try your luck with it in that sense (hilight gated works here for me), OR I would suggest using the density plot, which is better suited for per-object data. I realize the interactions with these visualizations are less-than ideal, BUT I’m happy to say that we are working hard to rewrite all of CPA in Python. When it is ready, we are looking forward to releasing a much more powerful, and easy-to-use analysis tool for image based screens! If you have any suggestions to help us guide our development, please do share it with us.

Thanks for using our software, and helping us improve it!
-Adam


#4

thanks! it is indeed working when using “per_image” information on the axes.