Skip to content

add midterm blog post for Soumyadip Niyogi (gsoc26)#1945

Open
captainvogon wants to merge 2 commits into
HSF:mainfrom
captainvogon:main
Open

add midterm blog post for Soumyadip Niyogi (gsoc26)#1945
captainvogon wants to merge 2 commits into
HSF:mainfrom
captainvogon:main

Conversation

@captainvogon

Copy link
Copy Markdown
Contributor

This pull request adds my midterm progress blog post for the GSoC 2026 project "Integration of CMS Combine with FCCAnalyses".

Copilot AI review requested due to automatic review settings July 17, 2026 10:02
@netlify

netlify Bot commented Jul 17, 2026

Copy link
Copy Markdown

Deploy Preview for earnest-hotteok-b1e1bf ready!

Name Link
🔨 Latest commit 820c959
🔍 Latest deploy log https://app.netlify.com/projects/earnest-hotteok-b1e1bf/deploys/6a5fd546857d940008112d1f
😎 Deploy Preview https://deploy-preview-1945--earnest-hotteok-b1e1bf.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify project configuration.

@github-actions github-actions Bot added the GSoC Related to Google Summer of Code activity label Jul 17, 2026

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This pull request adds a new GSoC 2026 midterm progress blog post to the HSF website content, describing the current state of the “Integration of CMS Combine with FCCAnalyses” project.

Changes:

  • Adds a new midterm blog post markdown file for Soumyadip Niyogi.
  • Documents current implementation status (datacard generation + execution flag) and a ROOT/Combine issue encountered during validation.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment on lines +1 to +6
| | |
| --- | --- |
| Name | [Soumyadip Niyogi](https://github.com/captainvogon) |
| Organisation | [IISER Thiruvananthapuram](https://www.iisertvm.ac.in/), [CERN](https://home.cern/), [HSF](https://hepsoftwarefoundation.org/) |
| Mentor | [Juraj Smiesko](https://kjvbrt.org/) (CERN), [Jan Eysermans](https://github.com/jeyserma) (MIT) |
| Project | [Integration of CMS Combine with FCCAnalyses](https://summerofcode.withgoogle.com/programs/2026/projects/FyDjzZ3Y) |
Initially, I thought our physics statistics were too low, or that the datacard was structurally broken. After some debugging with my mentors, we discovered the real culprit was ROOT 6.30.
The newest version of ROOT bundled in the Key4hep nightly builds has deprecated the old "silent boundary clipping" behavior. When running `AsymptoticLimits` on our perfectly smooth Asimov dataset, Combine's internal minimizer (Minuit) would aggressively guess a parameter boundary, step slightly out of bounds to test a value, and instantly trigger this new, strict ROOT exception, crashing the entire job!

**The Validation:** To bypass Combine's range-guessing and prove that my Python code was generating valid datacards, we switched the backend to run a `-M FitDiagnostics` test instead. The fit converged perfectly on our fake dataset, yielding `Best fit r: 1.0` with roughly a ~1.25% uncertainty. This proof that the datacards are structurally sound was a massive relief.
## Introduction
Hi again! We are at the midpoint of GSoC 2026, and I am excited to share the progress on my project bridging `FCCAnalyses` and `CMS Combine`. Since my introductory post, I have spent my time deep in the Python architecture of `FCCAnalyses`, transforming ROOT histograms from simulated e⁺e⁻ → ZH events into automated statistical models.

The primary goal of this first half was to build the core datacard generation engine. I am happy to report that the automated pipeline is now up and running. You can check out the finalized skeleton framework in [my latest Pull Request here](https://github.com/HEP-FCC/FCCAnalyses/pull/520).

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

it's no longer a skeleton :)

Over the past month, I have successfully implemented the `fit.py` interface within the `FCCAnalyses` framework. Here is what the tool can currently do:

* **Automated Datacard Generation:** The framework reads a user-defined, object-oriented Python configuration and automatically maps the requested signal and background processes to their respective ROOT shape histograms. It then spits out a perfectly formatted text-based `datacard.txt`.
* **The Execution Flag (`-e`):** To make the workflow completely seamless, I added a backend hook. By simply appending `-e` to the CLI command, the framework passes the newly generated datacard directly into a `subprocess` to execute `Combine` natively, eliminating the need for physicists to manually string together terminal commands.

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You can maybe mention that the additional parameters for Combine can be provided after -- or in the fit script.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

GSoC Related to Google Summer of Code activity

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants