Datasets:

ronantakizawa
/

github-top-developers

license: mit
task_categories:
  - text-classification
  - time-series-forecasting
  - text-retrieval
tags:
  - github
  - trending
  - developers
  - open-source
  - time-series
  - social-data
language:
  - en
size_categories:
  - 1K<n<10K
configs:
  - config_name: yearly
    data_files: yearly/data.csv
  - config_name: full
    data_files: full/data.csv

GitHub Top Developers by Year (2015-2025)

A derived dataset showing the top-ranked GitHub trending developers for each year, based on weighted scoring of their trending appearances across 41,841 raw data points from the Wayback Machine.

📊 Dataset Overview

Total Entries: 8,125 ranked developers
Years Covered: 2015 - 2025 (11 years)
Unique Developers: 4,763
Source: Derived from Wayback Machine snapshots of GitHub trending developers
Data Order: Sorted by year (descending: 2025 → 2015) and rank within each year
Update Frequency: Static historical dataset

🔧 Dataset Configurations

This dataset has two configurations defined in the YAML header:

Configuration: `yearly` (Default)

Top-ranked developers by year with 8,125 entries

from datasets import load_dataset
ds = load_dataset('ronantakizawa/github-top-developers', 'yearly')

Columns:

year (int): Year (2015-2025)
rank (int): Overall rank for that year (1 = highest score)
name (string): Developer/organization GitHub username
times_trended (int): Number of times appeared on trending
best_rank (int): Highest rank achieved (lowest number)
avg_rank (float): Average rank across all appearances
median_rank (int): Median rank
popular_repos (string): Top repositories (comma-separated)

Configuration: `full`

Complete daily trending data with 41,841 entries

from datasets import load_dataset
ds = load_dataset('ronantakizawa/github-top-developers', 'full')

Columns:

name (string): Developer/organization GitHub username
rank (int): Position in trending (1-25)
popular_repo (string): Associated repository at the time
date (string): Snapshot date (YYYY-MM-DD)

🏆 Scoring Methodology

Each developer's yearly score is calculated using:

Score = Σ (26 - rank) for each trending appearance

Where:
- Rank 1 = 25 points
- Rank 2 = 24 points
- ...
- Rank 25 = 1 point

Why this works:

✅ Rewards frequent appearances (more days trending = more points)
✅ Rewards high rankings (rank 1 is worth more than rank 25. We use 25 because github ranks the top 25 developers on their page)
✅ Balances consistency with peak performance

🌟 Key Insights

1. Year Winners (Highest Score Each Year)

Year	Winner	Score	Appearances	Notable Project
2025	comfyanonymous	257	18	ComfyUI
2024	emilk	2,052	124	egui (Rust GUI)
2023	hrydgard	1,858	111	PPSSPP emulator
2022	emilk	1,958	107	egui
2021	PySimpleGUI	1,993	120	PySimpleGUI
2020	stefanprodan	1,033	64	Flux CD
2019	Microsoft	308	15	Various
2018	google	325	15	Various
2017	facebook / Microsoft	77	4	(tie)
2016	facebook	485	23	React ecosystem
2015	facebook	48	2	React

Notable: emilk appeared on trending 124 times in 2024 alone (nearly every 3 days!)

2. All-Time Top 10 (Total Score Across All Years)

Rank	Developer	Total Score	Total Appearances	Years Active
1	emilk	6,311	370	2020-2025
2	hrydgard	5,181	324	2018-2024
3	stefanprodan	4,910	306	2018-2022
4	stephencelis	4,870	301	2016-2024
5	a8m	4,649	323	2016-2024
6	hathach	3,629	264	2018-2024
7	azure-sdk	3,621	251	2020-2024
8	twpayne	3,124	196	2017-2024
9	PySimpleGUI	3,059	185	2019-2023
10	arvidn	2,737	164	2017-2022

3. Trend Shifts Over Time

2015-2017: Organization Era

Big tech dominated: Facebook, Google, Microsoft
Individual developers rarely broke top 3
React ecosystem (Facebook) was the dominant force

2018-2019: Transition Period

Mix of organizations and influential individuals
Rise of open-source foundations (Apache, Linux Foundation)
Container/cloud technologies gained traction

2020-2025: Individual Developer Era

Individuals dominate top ranks consistently
emilk (egui) becomes most successful developer ever
Specialized tool creators rise (PySimpleGUI, hrydgard's PPSSPP)
AI/ML researchers become more prominent (rasbt, 2024-2025)

5. Notable Patterns

Extreme Consistency: emilk appeared 370 times across 6 years (average 62 times/year)
2021 Peak: PySimpleGUI set record with 120 appearances in a single year
Developer Longevity: sindresorhus maintained relevance from 2015-2021 (7 years)
Organization Decline: Big tech companies dropped from top spots after 2019
Ecosystem Impact: Most top developers maintain influential open-source libraries

💡 Usage Examples

Load with Hugging Face Datasets (Recommended)

from datasets import load_dataset

# Load yearly aggregated dataset (8,125 entries)
ds_yearly = load_dataset('ronantakizawa/github-top-developers', 'yearly')
df_yearly = ds_yearly['train'].to_pandas()

# Load complete daily dataset (41,841 entries)
ds_full = load_dataset('ronantakizawa/github-top-developers', 'full')
df_full = ds_full['train'].to_pandas()

# Get top 10 developers of 2024
top_2024 = df_yearly[df_yearly['year'] == 2024].head(10)
print(top_2024[['rank', 'name', 'times_trended', 'popular_repos']])

Time Series Analysis

import pandas as pd
import matplotlib.pyplot as plt
from datasets import load_dataset

ds = load_dataset('ronantakizawa/github-top-developers', 'full')
df = ds['train'].to_pandas()
df['date'] = pd.to_datetime(df['date'])

# Analyze a specific developer over time
developer = 'emilk'
dev_df = df[df['name'] == developer]

# Plot trending frequency over time
monthly_counts = dev_df.groupby(dev_df['date'].dt.to_period('M')).size()
monthly_counts.plot(title=f'{developer} Trending Frequency')
plt.ylabel('Days in Trending')
plt.show()

📖 Citation

If you use this dataset in your research, please cite:

@dataset{github_trending_developers_2015_2025,
  title={GitHub Top Developers Dataset (2015-2025)},
  author={Ronan Takizawa},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/ronantakizawa/github-top-developers}
}

📄 License

MIT License - Free to use with attribution

Last Updated: December 2025 Dataset Version: 1.0 Status: ✅ Complete and ready for use