Model Data Inventory

Shared model-data drops — contents, reproducibility, and EV / new-tech availability

Model Data Inventory — Shared Material (PC / RCA model + optimization tool)

Compiled: 2026-06-25 · Author of source material: Alon & Ishana (NZIPL modeling team) Purpose: inventory of two shared drops on OneDrive (000 Data Infrastructure/), what each contains, and — critically — whether the material is reproducible and whether EVs are covered.

1. `CICE Updated/` — the model inputs & per-tech RCA analyses

Folder holding the Predicted-Competitiveness (PC) / RCA model material.

Item	Size	What it is
`ML_vars.zip`	968 MB	Model variables + per-technology RCA analysis notebooks (see below)
`baci.zip`	3.07 GB	Raw CEPII BACI bilateral trade (the model’s trade input)
`Product Proximity Update.ipynb`	14 KB	Product-space proximity / relatedness update (standalone)

Inside `ML_vars.zip`

ML_vars/
├── Bentley_data_cleaning.R          # WDI download + cleaning (context variables)
├── wdi.xlsx · wdi_impute.csv        # World Bank Development Indicators (imputed) — model covariates
├── IRENA_INSPIRE_Patents_April_2024(INSPIRE_data).csv   # patent counts (feature)
├── Analysis/RCA <Tech> Analysis.ipynb   # 14 per-tech notebooks (the actual model runs)
└── RCA Construction/<Tech>/         # ✅ POPULATED — 146 files / 1.3 GB; per-tech variable-build
                                     #    files + PC values (see correction note below)

Correction (2026-06-25): an earlier version of this file said RCA Construction/<Tech>/ was empty. That was a tooling error on our side (a shell awk split on the space in “RCA Construction” and truncated the filenames). The model author confirmed — and a re-listing verified — that every tech folder is fully populated. Per the author: “ML vars has all the files that build the model variables; those are under RCA Construction; PC values are in each technology folder.”

Context / covariate data present (reproducibility inputs): - wdi_impute.csv + Bentley_data_cleaning.R — the WDI macro covariates (GDP, population, FDI, electricity use, governance, trade openness, manufacturing share, …) and how they’re built. - IRENA_INSPIRE_Patents…csv — patent feature. - baci.zip — raw trade (the RCA base).

Per-tech model notebooks (Analysis/) — 14 technologies, each a fully-executed notebook (RandomForest classifier on an RCA>1 target + shap.TreeExplainer for feature importance): Batteries · Biofuel · Electrolyzer · Geothermal · Heatpump · Magnets · Nuclear · Solar · Transmission · Wind · EV · plus DRI, DAC, Mass Timber (not in CVCE).

Reproducibility status — ✅ AVAILABLE

RCA Construction/ is fully populated — 146 files / 1.3 GB, with a subfolder per technology holding the variable-build inputs and the model’s PC outputs. Verified for all 14 techs: every folder carries PC scores, RCA-by-category, and the trade panel.

What	File (per tech, e.g. `EV/`)	Maps to
PC scores (country × year)	`<tech>_predicted_competition_all_years.csv`	`data/pc/pc_scores.parquet`
RCA by capability category	`<tech>_rca_by_category_year.csv`	`data/pc/pc_rca.parquet` (radar)
Product RCA (country × year × HS)	`hs17_rca_cyh.csv`	product-level RCA (PC scatter)
Trade panel	`<tech>_trade.csv`	RF feature build
Product-space proximity (EV only)	`ev_phi.csv`	relatedness/density

So the model is reproducible from this drop: the per-tech RCA-construction inputs are present (plus wdi_impute.csv, patents, baci.zip). The only piece not shipped as a tidy CSV is the per-category / per-HS6 SHAP table (the radar’s gold “need” polygon + the PC-scatter x-axis) — that is computed inside each Analysis/RCA <Tech> Analysis.ipynb (shap_by_cat, shap_df) and lifted from the notebook output rather than read from a file.

EVs specifically

RCA Construction/EV/ is the most complete folder (15–18 files), containing everything needed to wire EVs into the atlas:

ev_predicted_competition_all_years.csv (country_code, year, predicted_comp) → PC scores ✅
ev_rca_by_category_year.csv (country, year, category, …, RCA) → radar RCA (green) axis ✅
ev_phi.csv, ev_trade.csv, hs17_rca_cyh.csv (389 MB product RCA), BACI_HS17_V202601.zip (794 MB)
Notebooks: EV_Vars.ipynb, EV_Historical_Downscaling.ipynb; EVDataExplorer2025.xlsx

Task 3 status — unblocked. PC scores + RCA-by-category are ready to slot into data/pc/ for EVs (rebuilds the radar’s green polygon + PC-score readout). The radar’s gold SHAP polygon and the PC-scatter SHAP axis still need the per-category / per-HS6 SHAP, lifted from RCA EV Analysis.ipynb.

Three techs in the model not yet in CVCE

All fully populated with PC scores + RCA: DAC (direct air capture), DRI (direct reduced iron / green steel), Mass Timber. PC side is done; onboarding would need green-dict HS codes + a BACI extract (see CLAUDE.md “Adding a new technology”).

2. `Alon's optimization tool.zip` — CICE-V1 Interactive POC

A self-contained, email-able proof-of-concept dashboard: a country-competitiveness sensitivity tool. “If country X invested an extra ΔExport (as % of GDP) in technology T, how would its probability of being competitive — P(RCA>1) — move, and which HS categories drive it?”

web/
├── interactive_poc.html        # the dashboard (template + manifest + HS6 meta inlined)
├── interactive_template.html   # hand-written template (__INLINE_DATA__ substituted at build)
├── pipeline_flowchart.html + FLOWCHART_GUIDE.md   # methodology flowchart + its style guide
├── README.md                   # methodology + build instructions
└── data/
    ├── manifest.json           # {tech:{iso3:[years]}} availability + country names + ALPHAS grid
    ├── _hs_meta.js             # HS6 code → description cache
    └── <Tech>.json.gz          # one gzipped bundle per tech (country × year payloads)

Scope: 10 technologies (Batteries, Biofuel, Electrolyzer, Geothermal, Heatpump, Magnets, Nuclear, Solar, Transmission, Wind) × ~153 countries × 21 years (2003–2023). No EVs / DRI / DAC / Timber.

What the dashboard does (methodology): - Left panel — P(RCA>1) vs. investment expressed as ΔExport (% of GDP, with a $ axis). The blue curve is the full log-linear-form (LLF) model prediction; drag the marker to read P, ΔP, ΔExport (%GDP), ΔExport ($). An orange dotted “validity range” marks the largest investment α at which the local linearization stays within 50% of the full prediction. A static label reports P_baseline, marginal dP/dExport ($), and the validity threshold. - Right panel — RCA Impact (%) by HS2 category, recomputed live as you drag. Bar height = Σ_{i∈C} gᵢ² / Σ_all gⱼ² × 100 (bars sum to 100%); stacked segments are individual HS6 products by their L2² share. Categories: Chemicals, Industrial Materials, Electronics, Machinery, Metals, Textiles, Other.

Per-country-year payload (in each <Tech>.json.gz): pb (baseline P), gdp, exp (export level), val, and pf (the per-HS6 gradient / prediction vector) — the inputs that drive the draggable sensitivity curve.

Reproducibility: the bundles are build artifacts (gitignored upstream), regenerated by scripts/build_interactive_data.py --techs ALL --countries ALL --years ALL. That build script is not in this zip (only its outputs + the assembled HTML are). Runs over HTTP only (uses fetch); Plotly.js + pako load from CDN.

Bottom line

Question	Answer
Is there an EV model?	Yes — `RCA EV Analysis.ipynb`, fully run (RF + SHAP), May 15.
Is the EV data shipped?	Yes — `RCA Construction/EV/` is fully populated (15–18 files): PC scores, RCA-by-category, product RCA, trade, proximity, raw BACI.
Can we reproduce the model from this drop?	Yes — `RCA Construction/` holds the per-tech variable-build inputs + outputs for all 14 techs, plus WDI/patents/BACI. Only the tidy SHAP table isn’t a CSV (it’s in the notebooks).
Fastest route to EVs in the atlas	Slot `ev_predicted_competition_all_years.csv` → `pc_scores` and `ev_rca_by_category_year.csv` → `pc_rca`; lift SHAP-by-category from `RCA EV Analysis.ipynb` for the radar gold polygon / scatter x-axis.
Techs in the model but not in CVCE	DAC, DRI, Mass Timber — PC side done; need green-dict + BACI extract to onboard.
Does Alon’s tool cover EVs?	No — 10 techs, EVs not among them.