cytomining / cytomining/CytoTable

How do we determine what CellProfiler `features` are `metadata` or not?

Open
#294 1 comment 1 reaction 0 assignees View on GitHub
question
Dominant language
Python
Stars
21
Forks
6
Avg merge
2d 3h
Merged PRs (30d)
6

Description

# How do we determine what CellProfiler `features` are `metadata` or not?

Hello! I am opening up this issue to start a conversation on how CytoTable should handle some columns we get as CellProfiler outputs.

Specifically, I noticed that in the most recent version of CytoTable at the time of this issue, I was seeing columns called "Nuclei_Parent," "Nuclei_Child," and "Nuclei_Number" that were being included as features after rerunning.

From my perspective, this column should not be treated as features; they are mainly important for merging single cells.

There are other features like center x,y coordinates that are included as `features` and have to be manually updated as `metadata`. There is a lot of nuance to this and likely a good internal conversation to have and make some main decisions:

1. Should CytoTable be the filter for what is considered features or metadata?
2. What columns do we determine are `metadata` or `features`?
3. Dicated by the above, should we set these decisions as standards or optional parameters?

Contributor guide

Open the contributing guide

Research direction

No files or tests are named. Begin with the CellProfiler-output handling path and inspect how Nuclei_Parent, Nuclei_Child, Nuclei_Number, and center coordinates are classified after reruns. Done requires an agreed distinction between metadata and features, with the implementation scope and any optional parameters or tests defined.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
data
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.