Multiple tables

Hi again,

I am trying to figure out how to process a quite large amount of images using CPA. The problem I am having comes from the necessity of going through multiple run of CellProfiler (due in part to memory limitations…) and so having multiple output files. I’d like either to be able of processes them all as part of a single experiment or to be able to save (export) the rules and then reload them with a new set of images.

Thanks for any help,



Very glad you tried the forum for this one because we have a wizard written in python that will get you where you need to be.

Here’s how it works:

  • The wizard looks for pairs of image & object tables and presents them to you in a list.

  • You select all the table-pairs you wish to include in your analysis.

  • It then merges all the image tables into a separate master image table, and all the object tables into a separate master object table.

  • The image and object numbers are preserved in all tables, the wizard simply adds a “TableNumber” column which you must specify in your properties file by “table_id=TableNumber”

Note: if you create new tables later that you wish to add to your analysis, you will need to repeat this step with the original tables + the new table.

The script currently requires python 2.5, and the following 3 packages: wxPython, MySQLdb, numpy.

I have just finished the new distribution of CPA 2.0 along with an accompanying manual, and I am in the process of figuring just HOW to include this tool. I will email you the script as it is in case you need it sooner rather than later – I don’t recall whether you are running CPA from source or from a distribution.

If you’re running from source, then you should be pretty much set to use this – in fact, it is under src/ Otherwise, let me know and I’ll help you get set up with the necessary prereqs.


Oye, I should have mentioned that the wizard does not currently function for SQLite databases, so let me know if this is your situation. I’ll be working on a way to handle these in the meantime.