[GSoC Project Proposal]: phytoclass for the masses

[7yl4r]

Project Description

Researchers collect water-pigment data using High-Performance Liquid Chromatography (HPLC) throughout the globe. Among the uses of this data is the ability to estimate phytoplankton abundance. A powerful program for this analysis has been CHEMTAX, but recently a new R package, phytoclass, has emerged with improved functionality.

This project will improve usability and accuracy of phytoclass.

Expected Outcomes

Suggested improvements to the phytoclass ecosystem include:

• better error checking & messaging in the phytoclass R package
• continued development on the phytoclass R-Shiny GUI
• enhancement of phytoclass functionality to include regional variations for phytoplankton estimation

Skills Required

R. R+Shiny.

Additional Background/Issues

• phytoclass source code
• phytoclass publication
• phytoclass R-Shiny GUI
• CHEMTAX

Mentor(s)

@7yl4r, @alexanderhayward1995

Mentor Contact Email(s)

[email protected]

Expected Project Size

90 hours

Project Difficulty

Intermediate

[mwengren]

@7yl4r Thanks! Could you submit a PR against our ideas list to officially add these new projects: https://github.com/ioos/gsoc/blob/main/2025/ideas-list.md.

The GSoC review may have already happened, but it can't hurt to add a few more ideas before they officially announce accepted orgs on 2/27.

[7yl4r]

@mwengren : Oops. I didn't realize I missed the deadline. Here's a PR for the two project ideas I added : #92

[celestinelindarto]

Hi @7yl4r ,

Thank you for organizing this project and providing an opportunity to contribute via GSoC. I’m interested in improving phytoclass and had a few questions:

• The description mentions that phytoclass has improved functionality over CHEMTAX - could you please elaborate on what makes the package unique?
• What are some of the most common errors or issues that users encounter when using phytoclass?
• Could you provide more details on the improvements needed for the phytoclass R-Shiny GUI?
• How does the regional variation for phytoplankton estimation work? What specific functionality needs to be developed?
• I noticed several open issues in the phytoclass repository - is there any specific task that you would like me to work on first?

I look forward to hearing from you. Thank you for your time.

[7yl4r]

* The description mentions that phytoclass has improved functionality over CHEMTAX - could you please elaborate on what makes the package unique?

Great question! I've added a section to the README to address this.

* What are some of the most common errors or issues that users encounter when using phytoclass?

The open issues on the phytoclass repo list some of the errors, but some of the work in this project will be identifying errors by testing with different datasets.

* Could you provide more details on the improvements needed for the phytoclass R-Shiny GUI?

There is a monthly phytoclass working group (with 3 regular attendees) that are working to figure this out.
You would be expected to attend and directly interface with users to interpret their needs.

* How does the regional variation for phytoplankton estimation work? What specific functionality needs to be developed?

Different regions have different phytoplankton species. Adding in this information will improve estimation accuracy significantly. phytoclass's current customization features can do this, but it is difficult. The goal would be to make this feature easier to use by modifying the API functions.

* I noticed several open issues in the phytoclass repository - is there any specific task that you would like me to work on first?

The issues listed in the repository are stale. First steps of this project will be narrowing scope and creating some issues.

It is not expected that all of the things above will be addressed. Choose one or two that are most interesting to you. I'm happy to answer more questions here or over email ([email protected])

[celestinelindarto]

Thank you for your reply @7yl4r. I will communicate with you via email.