Author: Raimundo Elias Gomez Affiliations: CONICET / National University of Misiones (Argentina); Faculty of Arts, University of Porto (Portugal) Contact: elias.gomez@conicet.gov.ar ORCID: 0000-0002-4468-9618
This repository contains the data, analysis scripts, and figures for the article "The spatiality of software: subnational economic complexity from GitHub data in Argentina", currently under peer review.
The study constructs an Economic Complexity Index for software production (ECIsoftware) at the level of 224 Argentine departments using a bipartite network of departments and 87 programming languages derived from 229,270 geocoded GitHub repositories. A three-stage analytical strategy — Multiple Correspondence Analysis (MCA), Hierarchical Agglomerative Clustering (CAH), and type-specific regressions — examines how the determinants of software complexity vary across six territorial types.
github-subir/
├── README.md
├── data/ # Processed datasets and summary tables
│ ├── departments_full.csv # All 511 departments: MCA coords, clusters, ECI, census vars
│ ├── bipartite_matrix.csv # 224 depts x 87 languages (repo counts, filtered)
│ ├── rca_binary_matrix.csv # 224 x 87 binary RCA matrix (threshold >= 1)
│ ├── eci_ranking_FINAL.csv # ECI ranking for 224 departments
│ ├── table_01_eci_ranking_full.csv # ECI ranking with sociodemographic variables
│ ├── table_02_pci_ranking_languages.csv # PCI ranking for 87 programming languages
│ ├── table_03_cluster_profiles.csv # Mean profiles of 6 departmental types
│ ├── table_04_regression_summary.csv # Regression coefficients by type
│ ├── table_05_key_numbers.csv # Summary statistics (key-value)
│ ├── table_06_crossvalidation_geo.csv # Geospatial cross-validation (511 depts)
│ └── regression_output_FINAL.txt # Full regression output (text)
├── figures/ # Article figures (300 DPI)
│ ├── fig_01_pci_ubiquity.png # Figure 1: PCI vs ubiquity (87 languages)
│ ├── fig_02_mca_biplot.png # Figure 2: MCA biplot (Axes 1-2, N=511)
│ ├── fig_03_cah_mca_clusters.png # Figure 3: Six types in MCA space
│ ├── fig_04_cluster_maps.png # Figure 4: Spatial distribution of types
│ ├── fig_05_eci_vs_devs.png # Figure 5: ECI vs developer density
│ ├── fig_06_forest_plot.png # Figure 6: Forest plot of betas by type
│ ├── fig_S1_dendrogram.png # Figure S1: Ward's dendrogram (k=6)
│ └── fig_S2_diagnostics_panel.png # Figure S2: MCA scree + cluster quality
├── scripts/ # Analysis pipeline (Python)
│ ├── 00_build_schema.py # Stage 0: Integrate 11 data sources into art1 schema
│ ├── 01_compute_eci.py # Stage 1: Compute ECI via eigenvalue decomposition
│ ├── 02_mca.py # Stage 2a: Multiple Correspondence Analysis (8 vars, N=511)
│ ├── 03_cah.py # Stage 2b: Ward's CAH on MCA coordinates (k=6)
│ ├── 04_regressions_by_type.py # Stage 3: Pooled + type-specific regressions, Chow test
│ ├── 05_regenerate_figures.py # Generate all 8 figures (6 article + 2 supplementary)
│ └── 06_cluster_maps.py # Generate Figure 4 (3x2 small-multiples map)
├── audit/ # Data quality and geocoding validation
│ ├── audit_01_full_province_department.csv # Raw vs geo-validated counts (513 depts)
│ ├── audit_02_discrepancies.csv # 32 departments with discrepancies
│ ├── audit_03_province_summary.csv # Province-level data integrity summary
│ ├── audit_04_foreign_users.csv # 76 excluded non-Argentine users
│ ├── audit_05_foreign_repos_by_dept.csv # Departments affected by foreign repos
│ ├── audit_06_ambiguous_users_sample.csv # 31 ambiguous location samples
│ └── audit_07_eci_before_after.csv # ECI ranking before/after corrections
└── supplementary/ # Supplementary material
├── supplementary_tables.md # Supplementary tables and figures
├── table_S1_eci_full_ranking.csv # Full ECI ranking (224 departments)
├── table_S2_cluster_region_crosstab.csv # Cluster × region cross-tabulation
├── table_S3_small_types_data.csv # Data for small-N types (Peripheral, Semi-Rural)
└── table_S4_within_type_correlations.csv # Within-type correlations with ECI
| File | Rows | Columns | Description |
|---|---|---|---|
departments_full.csv |
511 | 28 | All Argentine departments with census (2010), MCA coordinates (5 dims), cluster assignment, ECI, GitHub metrics |
bipartite_matrix.csv |
224 | 88 | Repository counts by department and programming language (dpto5 + 87 languages) |
rca_binary_matrix.csv |
224 | 88 | Binarised Revealed Comparative Advantage (RCA >= 1) |
table_02_pci_ranking_languages.csv |
87 | 5 | Product Complexity Index for programming languages |
| Variable | Source | Description |
|---|---|---|
dpto5 |
INDEC | Five-digit department code |
region |
Derived | Six regions: CABA, Pampeana, NOA, NEA, Cuyo, Patagonia |
pob_2010, pob_2022 |
Census | Population |
pct_jefe_sec_2010 |
Census 2010 | % household heads with secondary education |
pct_jefe_uni_2010 |
Census 2010 | % household heads with university education |
pct_pc_2010 |
Census 2010 | % households with computer |
pct_nbi_2010 |
Census 2010 | % with unsatisfied basic needs (poverty) |
pct_hacinam_2010 |
Census 2010 | % overcrowding |
rad_2014 |
VIIRS | Mean nighttime radiance (2014) |
tasa_empleo_2010 |
Census 2010 | Employment rate |
mca_dim1...mca_dim5 |
MCA | Factorial coordinates (5 retained axes) |
mca_cluster |
CAH | Cluster number (1-6) |
mca_cluster_label |
CAH | Cluster label |
eci_software |
ECI | Economic Complexity Index (standardised) |
eci_diversity |
ECI | Number of languages with RCA >= 1 |
eci_avg_ubiquity |
ECI | Mean ubiquity of RCA languages |
gh_total_developers |
GitHub | Total geocoded developers |
gh_total_repos |
GitHub | Total repositories |
gh_devs_per_10k |
Derived | Developers per 10,000 inhabitants |
gh_hill_q1_shannon |
GitHub | Language diversity (Shannon entropy) |
The scripts are numbered in execution order and depend on a PostgreSQL database (posadas) with the source data. The pipeline proceeds as follows:
-
00_build_schema.py— Integrates 11 data sources (Census 2010/2022, VIIRS nighttime lights, NDVI, GitHub, ENACOM) into a single analysis-ready table (art1.departamentos, 511 departments, ~208 columns). -
01_compute_eci.py— Constructs the bipartite network (departments x languages), computes RCA, and extracts ECI and PCI via eigenvalue decomposition of the normalised adjacency matrix. Applies geocoding corrections (Cordoba shift, CABA aggregation, foreign user exclusion). -
02_mca.py— Multiple Correspondence Analysis on 8 pre-treatment variables discretised into terciles (24 modalities, N=511). Retains 5 axes via Benzecri correction. Projects ECI and developer metrics as supplementary variables. -
03_cah.py— Ward's hierarchical clustering on 5 MCA coordinates. Selects k=6 (silhouette=0.330, Calinski-Harabasz=224.5). Profiles clusters with ANOVA and chi-squared tests. -
04_regressions_by_type.py— Pooled and type-specific OLS regressions of ECI on pre-treatment predictors. Chow test for structural heterogeneity. Forest plot of standardised coefficients. -
05_regenerate_figures.py— Generates all 8 figures (6 article + 2 supplementary) with unified formatting (300 DPI). -
06_cluster_maps.py— Generates Figure 4 (3x2 small-multiples map of cluster spatial distribution) using PostGIS geometries.
- ECIsoftware is distinct from developer counts: r = 0.47 (moderate correlation)
- PCI validates the framework: scientific computing languages (Erlang, Fortran, Julia) rank as most complex; web technologies (JavaScript, HTML, CSS) as least complex
- Six departmental types explain 30.2% of ECI variance (eta-squared = 0.302)
- Determinants are structurally heterogeneous: education drives complexity in Metropolitan-Core; computer ownership in Metropolitan-Diversified; population alone in Pampeana-Educated; no predictor significant in Intermediate-Urban
| Source | Period | Coverage | Access |
|---|---|---|---|
| GitHub API | Accumulated through 2025 | 229,270 repos, 23,619 users | Scraped early 2026 |
| Census (INDEC) | 2010, 2022 | 511 departments | datos.gob.ar |
| VIIRS DNB | 2014 | Department-level radiance | Google Earth Engine |
| ENACOM | ~2023 | Internet infrastructure | datosabiertos.enacom.gob.ar |
python >= 3.10
numpy
pandas
scipy
scikit-learn
prince
matplotlib
seaborn
geopandas
sqlalchemy
psycopg2
If you use these data or methods, please cite:
Gomez, R. E. (2026). The spatiality of software: subnational economic complexity from GitHub data in Argentina. Working paper.
Zenodo DOI:
Data and code are provided under the CC BY 4.0 licence.