Datasets:

hreyulog
/

GitHub-3Repo-7User-Opinion-Dynamics

repository stringclasses 3 values	pr_time stringdate 2017-02-01 00:00:00 2021-03-01 00:00:00	dev_01 float64 0.11 0.53	dev_02 float64 0.07 0.67	dev_03 float64 0.05 0.52	dev_04 float64 0.05 0.51	dev_05 float64 0.08 0.64	dev_06 float64 0.09 0.42	dev_07 float64 0.04 0.55
ceph	2017-02-01	0.405821	0.54148	0.33737	0.325392	0.447851	0.406066	0.31766
ceph	2017-03-01	0.317623	0.318124	0.391112	0.311146	0.401466	0.335026	0.299413
ceph	2017-04-01	0.358813	0.476688	0.384133	0.36973	0.310626	0.378038	0.227391
ceph	2017-05-01	0.316523	0.444118	0.324935	0.309026	0.307694	0.369036	0.541795
ceph	2017-06-01	0.274533	0.427552	0.349383	0.216233	0.257445	0.334827	0.492357
ceph	2017-07-01	0.533073	0.333774	0.399655	0.30938	0.351704	0.342256	0.309128
ceph	2017-08-01	0.334629	0.304565	0.522204	0.332381	0.253609	0.367453	0.307529
ceph	2017-09-01	0.408014	0.338392	0.372152	0.340914	0.332969	0.402054	0.258176
ceph	2017-10-01	0.342018	0.33867	0.369391	0.448808	0.381982	0.328663	0.316098
ceph	2017-11-01	0.308258	0.446121	0.366899	0.32965	0.367038	0.418106	0.316019
ceph	2017-12-01	0.30284	0.317553	0.456725	0.50619	0.323387	0.321851	0.336119
ceph	2018-01-01	0.380767	0.357885	0.340052	0.503699	0.418204	0.339189	0.330222
ceph	2018-02-01	0.333409	0.378046	0.347105	0.334108	0.311408	0.319224	0.309018
pytorch	2020-03-01	0.118869	0.113872	0.109554	0.141384	0.163167	0.128435	0.232424
pytorch	2020-04-01	0.112541	0.142278	0.092828	0.121792	0.222753	0.139873	0.100723
pytorch	2020-05-01	0.107804	0.133014	0.09164	0.123145	0.105387	0.107956	0.050834
pytorch	2020-06-01	0.12204	0.113278	0.102521	0.134523	0.117094	0.189724	0.131301
pytorch	2020-07-01	0.135381	0.109286	0.105097	0.14985	0.084588	0.190605	0.087531
pytorch	2020-08-01	0.332659	0.172441	0.091088	0.142246	0.334363	0.114015	0.043995
pytorch	2020-09-01	0.189711	0.200149	0.131224	0.105976	0.104595	0.151496	0.248198
pytorch	2020-10-01	0.158887	0.092435	0.13737	0.106111	0.088096	0.122398	0.196578
pytorch	2020-11-01	0.275277	0.131316	0.113544	0.087541	0.09457	0.142492	0.102076
pytorch	2020-12-01	0.208208	0.108294	0.09979	0.101425	0.123864	0.109858	0.258042
pytorch	2021-01-01	0.134311	0.113248	0.115409	0.114501	0.191008	0.111342	0.315964
pytorch	2021-02-01	0.174074	0.138918	0.123549	0.147504	0.085437	0.094433	0.095393
pytorch	2021-03-01	0.175217	0.105569	0.090351	0.159	0.086034	0.106812	0.21941
swift	2019-04-01	0.272151	0.210707	0.119871	0.163719	0.197534	0.185559	0.216851
swift	2019-05-01	0.33572	0.20415	0.13244	0.140887	0.295342	0.253218	0.335158
swift	2019-06-01	0.259662	0.491992	0.116131	0.179663	0.63671	0.210051	0.255107
swift	2019-07-01	0.359392	0.122497	0.120815	0.162293	0.202171	0.203706	0.072542
swift	2019-08-01	0.215401	0.145039	0.16934	0.053438	0.155883	0.135283	0.24377
swift	2019-09-01	0.421057	0.665907	0.099077	0.149355	0.211051	0.209274	0.22067
swift	2019-10-01	0.260789	0.31177	0.112	0.201537	0.170421	0.099361	0.114454
swift	2019-11-01	0.325617	0.067258	0.133813	0.252079	0.276531	0.145727	0.196205
swift	2019-12-01	0.311961	0.281502	0.12687	0.197168	0.248248	0.116745	0.179869
swift	2020-01-01	0.108178	0.18063	0.054831	0.211054	0.1636	0.208359	0.553246
swift	2020-02-01	0.117053	0.203047	0.095594	0.188065	0.308223	0.187961	0.067309
swift	2020-03-01	0.210832	0.177261	0.131544	0.184723	0.279933	0.16076	0.396723
swift	2020-04-01	0.211267	0.252058	0.103138	0.368869	0.340563	0.204194	0.444471

GitHub 3Repo 7User Opinion Dynamics

This dataset contains monthly opinion-dynamics time series derived from three large open-source GitHub repositories: Ceph, PyTorch, and Swift. Each CSV file represents one repository and contains a repository label, one timestamp column, and seven anonymized developer trajectory columns.

Files

file	rows	anonymized developer columns
`ceph.csv`	13	7
`pytorch.csv`	13	7
`swift.csv`	13	7

Schema

Each CSV uses the same schema:

repository: source repository label (ceph, pytorch, or swift).
pr_time: monthly timestamp in YYYY-MM-DD format.
dev_01 ... dev_07: anonymized opinion-dynamics values for the seven selected developers in that repository.

Developer identifiers from the source analysis are intentionally not included. The anonymized column order is stable within each CSV, so longitudinal analysis remains reproducible, but the public dataset does not expose the original GitHub handles.

Usage

from datasets import load_dataset

ds = load_dataset("hreyulog/GitHub-3Repo-7User-Opinion-Dynamics")
print(ds)

You can also load an individual CSV with pandas:

import pandas as pd

ceph = pd.read_csv("ceph.csv")

Citation

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

@article{HE2026102824,
title = {Social life of code: Modeling evolution through code embedding and opinion dynamics},
journal = {Journal of Computational Science},
volume = {96},
pages = {102824},
year = {2026},
issn = {1877-7503},
doi = {https://doi.org/10.1016/j.jocs.2026.102824},
url = {https://www.sciencedirect.com/science/article/pii/S1877750326000426},
author = {Yulong He and Nikita Verbin and Sergey Kovalchuk},
keywords = {Opinion dynamic, NLP, Human behavior analysis, Codebase evolution, Social-technical analysis}
}

Paper

Downloads last month: 30

Number of rows:

Total file size:

11.1 kB