enh: proceedings

manuscript/main.tex

@@ -0,0 +1,70 @@
+\documentclass{article}
+\usepackage{markdown}
+\usepackage{graphicx}
+
+\title{Proceedings of the OHBM Hackathon 2023}
+\author{Yu-Fang Yang, Anibal Sólon Heinsfeld}
+\date{December 2023}
+
+\begin{document}
+
+\maketitle
+
+\section{Introduction}
+
+Following the previous editions, the Organization for Human Brain Mapping (OHBM) 2023 embraced a hybrid format, underscoring our commitment to diversity, inclusion, and equitable access.
+This approach is also aimed at reducing the carbon footprint and addressing barriers such as visa issues, enabling broader participation from the global neuroscience community.
+The OHBM Brainhack is a hackathon to the main OHBM meeting, where neuroscientists of various expertise levels and academic interests convene to collaborate in a dynamic, hackathon-style environment.
+These collaborations focus on shared interests from any participant and foster a culture of open science.
+The OHBM Brainhack 2023 was rooted in three core principles.
+First, it aimed to provide an inclusive collaborative space for interdisciplinary interaction among scientists at all career stages.
+This aspect was embodied in the “Hack-Track,” where participants worked together to advance significant projects needing community support.
+Second, the event emphasised empowering scientists to enhance their research quality through hands-on training in open-science practices, a goal realised in the “Train-Track.” Lastly, introducing a buddy system was a novel feature for this edition of the OHBM Brainhack.
+This system paired newcomers with experienced mentors, offering guidance and support to first-time participants, ensuring a welcoming and enriching experience for all.
+In the following sections, we delve into the specifics of the Hack-Track and Train-Track of OHBM 2023 Brainhack, detailing the proceedings and highlighting the key outcomes and advancements made during the event.
+
+\subsection{Train-Track}
+
+The 2023 edition of the OHBM Brainhack continues its educational mission with the Train-Track, offering a diverse range of fundamental courses such as “Introduction to Jupyter Notebook” and “Basics of Writing Good Code”, along with cutting-edge topics like “Container for Science” and “Data Management with Datalad.”
+This initiative enriches participants' learning experiences by fostering a study group-like environment, augmented by our buddy system that promotes peer-to-peer learning and collaboration.
+Additionally, we introduced onsite sign-ups for mentor-mentee pairings in specific topics, further encouraging group formation among attendees with shared learning goals and facilitating networking opportunities.
+
+The collaborative setting of Train-Track allowed participants to engage with like-minded peers, creating a supportive space for knowledge exchange and skill development.
+Course topics were carefully selected based on submitted projects, primarily utilising Python, Git, and GitHub.
+We also expanded topics to include “Reproducible Workflows of The Turing Way”, “Transparent MRI”, and “Machine Learning for Neuroimaging and Reproducibility Checklist”.
+Notably, these courses were pre recorded and made available on YouTube by the OHBM Open Science Interest Group (OSSIG), ensuring accessible learning resources.
+Participants were encouraged to use various communication platforms, including in-person interactions and digital channels like Discord, to foster this collaborative learning process.
+The Train-Track segment at OHBM Brainhack 2023 truly embodied the spirit of community and collective growth.
+
+
+\subsection{Buddy System}
+
+To involve newcomers in the Hackathon, we created the Buddy System.
+It pairs the participants in a mentor-mentee fashion, allowing mentors to help mentees throughout the Hackathon, giving guidance regarding interests, projects to join, and networking.
+This system was set up in such a way that the pairs were assigned at random, avoiding creating any biases in the matching.
+
+\subsection{Hack-Track}
+
+Before the OHBM Hackathon 2023 commenced, project leaders submitted their proposals to our repository, based on Github, setting the stage for a dynamic and inclusive event.
+At the start of the Brainhack, these leaders had the opportunity to pitch their projects to the audience.
+During these presentations, each leader was allotted a few minutes to outline their project's goals, current status, and aspirations for the hackathon and its future trajectory.
+A comprehensive list of these diverse and innovative projects can be found at our main website \footnote{https://ohbm.github.io/hackathon2023/hacktrack/}.
+The range of topics and scopes of these projects was impressively varied, encompassing areas such as coding, writing, documentation, community guideline development, and visualisation techniques.
+This extensive range allowed participants to tailor their involvement to their interests and goals.
+Some dedicated their efforts to a single project, immersing themselves deeply in a specific area or task.
+Others divided their time among multiple projects, seeking a broader experience.
+Participants selected projects based on various factors.
+For some, it was an opportunity to contribute to a field in which they already had expertise.
+Others saw it as a chance to acquire new skills or knowledge in an unfamiliar area.
+
+Additionally, the Hackathon projects often introduced novel methods or programming languages, offering participants practical skills that could be applied to future endeavours.
+Beyond the technical and educational aspects, the event also fostered community building; many were drawn to projects simply for the opportunity to collaborate with and learn from a new group of peers.
+Over three intensive days, the 2023 edition of the OHBM Hackathon successfully propelled forward [XX – number of projects] projects, thanks to the collaborative efforts of [xx – number of scientists] scientists hailing from countries across Africa, Asia-Pacific, Europe, the Middle East, North America, and South America.
+Despite the inherent challenges of a hybrid event, especially in synchronising video streams, we effectively utilised various technologies within a single platform, Discord, to facilitate seamless participant exchanges.
+This approach ensured that irrespective of geographical location, every participant could engage fully in the hackathon's collaborative spirit.
+
+\subsection{Projects}
+
+\markdownInput{projects.md}
+
+\end{document}

scripts/fetch_gh_proceedings.py

@@ -0,0 +1,113 @@
+#!/bin/env python
+import os
+import re
+import requests
+import yaml
+import json
+
+def fetch_gh_issues():
+
+    GH_AUTH = os.environ['GH_AUTH']
+    REPO = 'ohbm/hackathon2023'
+    ISSUE_LABEL = 'Proceedings'
+    # ISSUE_READY_LABEL = 'Good to go'
+    ISSUE_FILTER = f'labels={ISSUE_LABEL}&per_page=100'
+    URL = f'https://{GH_AUTH}@api.github.com/repos/{REPO}/issues?{ISSUE_FILTER}'
+
+    with open('.github/ISSUE_TEMPLATE/hackathon-proceedings.yml') as f:
+        issue_form = yaml.safe_load(f)
+
+    fields = issue_form['body']
+    fields = [f for f in fields if f['type'] != 'markdown']
+
+    res = requests.get(URL)
+    issues = res.json()
+    issues_list = []
+    for issue in issues:
+
+        print("looking for issue", issue['url'])
+        # if ISSUE_READY_LABEL not in [i['name'] for i in issue["labels"]]:
+        #     continue
+
+        if issue["state"] != "open":
+            continue
+
+        print('passed')
+
+        try:
+
+            body = issue["body"]
+            lines = [l.strip() for l in body.replace('\r\n', '\n').split('\n')]
+
+            field_ordering = []
+            for field in fields:
+                field_start = None
+                field_label = field['attributes']['label']
+
+                for li, line in enumerate(lines):
+                    is_line_title = line.startswith(f'### {field_label}')
+                    if field_start is None and is_line_title:
+                        field_start = li
+
+                field_ordering += [(field, field_start)]
+
+            field_ordering = list(sorted(field_ordering, key=lambda f: f[1]))
+
+            issue_info = {}
+
+            field_bounds = zip(field_ordering, field_ordering[1:] + [(None, None)])
+            for (field, i), (_, ni) in field_bounds:
+                field_id = field['id']
+                field_label = field['attributes']['label']
+
+                if i is None:
+                    issue_info[field_id] = None
+                    continue
+
+                field_value = '\n'.join(filter(None, lines[i+1:ni]))
+                field_value = re.sub(
+                    r'<!--.*?-->', '', field_value,
+                    flags=re.DOTALL
+                )
+                field_value = field_value.strip()
+
+                if field_value == '_No response_':
+                    field_value = None
+
+                if field['type'] == 'checkboxes':
+                    field_options_labels = [
+                        o['label'].strip()
+                        for o in field['attributes']['options']
+                    ]
+                    field_selected_options = []
+                    field_options_value = field_value.split('\n')
+                    for l in field_options_value:
+                        if l[6:] not in field_options_labels:
+                            continue
+                        if l.startswith('- [X] '):
+                            field_selected_options.append(l[6:])
+                        if l.startswith('- [x] '):
+                            field_selected_options.append(l[6:])
+
+                    field_value = field_selected_options
+
+                issue_info[field_id] = field_value
+
+            issue_info['issue_link'] = issue["html_url"]
+            issue_info['issue_number'] = issue["number"]
+            issues_list.append(issue_info)
+        except Exception as e:
+
+            print("error in issue", issue['url'])
+            print(e)
+            raise e
+
+    with open('./public/proceedings.json', 'w') as f:
+        json.dump(issues_list, f, indent=2)
+
+
+if __name__ == '__main__':
+    from dotenv import load_dotenv
+    load_dotenv()
+
+    fetch_gh_issues()