41-alchemy-data-three

Every script lives inside the blueprints/ directory of the mirrored v2 engine. Together they form the complete analysis, generation, and access stack.

Three additional .py files (totalling 18) are present in the assembled package's data-access, core-systems, or utility handoff directories; they provide the original v1 probability library (alchemy_probability_library), the multi-degree probability engine (alchemy_multi_degree_probability), and supporting helpers that alchemy_data_v2.py imports.

The CSV tables are the distilled, human-readable output of the analysis engine. Each row represents one element combination scored across nine dimensions (see "Probability matrices" below). The tables are organised into four compound categories — metals (alloys and intermetallics), solvents (organic, protic, and ammoniacal), gases (single-phase mixtures and noble-gas blends), and functional objects (exotic multi-element compounds) — and four scoring slices.

Four additional CSVs — value_metals.csv, value_solvents.csv, value_gass.csv, value_objects.csv — rank every combination by composite value score, segmented into top, middle, and bottom slices. Each row carries the full nine-column property vector. For example, value_metals.csv shows that AlFeSi (ternary intermetallic, value 17.66, Mohs 7.0, $1/kg, availability 9.75) leads the metals, while quaternary intermetallics like AlFeMgSi (value 17.15) and FeMgSi (value 16.82) follow.

The remaining 18 CSVs in the full package are generated outputs, including the 96-333-fixed.csv master table and additional per-chunk exports that the find_novel_outputs.py script produces during full data sweeps.

The bluray_data/ directory is the heart of the system — a self-contained probability database designed to fill a 25 GB Blu-ray disc. It contains 2,609 files (2,605 gzip-compressed JSON archives plus four plain-text/JSON metadata files).

The SVG collection breaks down into seven functional categories.

The remaining SVGs in the full 195-file count are additional generated compound diagrams and patent-figure variants produced during extended data sweeps.

A WordPress export file (alchemy-blog-posts.wxr.xml, WXR format, 20 KB) contains four long-form articles intended as companion documentation for the system:

The 41-41-41/ folder is the third-generation layer that transforms the v2 data engine into a legally filed invention. It contains four files totalling approximately 5.8 MB:

At the package root, N417.pdf provides the USPTO N-417 filing receipt — legal evidence that the application was received and docketed. A compressed copy of the subtree is also provided as 41-41-41.zip.

In short, v2 is the engine; the 41-41-41 subtree is the patent filing that turns the engine into filed intellectual property.

The Alchemy Data substrate underlies three disease-discovery patent applications in the inventor's portfolio:

All three applications depend on the compound-discovery pipeline, the probability matrices, and the transformation database that ship with this package. Acquiring Alchemy Data Three means acquiring the substrate that powers these filings.

The system's analysis pipeline operates in five stages:

Every compound combination in the system is scored across nine dimensions, each computed from per-element property tables built into the find_novel_outputs.py script:

The ElementTransformationDatabase stores transformations as structured records. Each transformation has:

The FormulaSynthesisMapper extends this with complete invention recipes: given a formula, it outputs named synthesis methods with temperature, equipment, multi-step instructions, and success probabilities. The system currently maps nine physics/chemistry formulas (E = mc², E = hf, PV = nRT, F = ma, and others) to concrete synthesis routes. The method_catalog_analyzer enumerates 257,500+ base known methods (772,500 with condition variations) across categories: enzymatic reactions (100,000), named organic reactions (50,000), catalytic reactions (30,000), inorganic reactions (20,000), materials synthesis (15,000), polymerisation (10,000), electrochemical (8,000), and more.

The 5,210 .gz files are gzip-compressed JSON documents. Each preserves a complete, self-describing data snapshot:

All archives use Python's gzip and json modules for read/write. The BluRayDataReader class handles transparent decompression and caching when querying the corpus.

Run python bluray_data_api.py (or double-click START_WEB_INTERFACE.bat) to start the Flask REST API on localhost. Then open bluray_data_web.html in a browser. The tabbed interface provides compound search, transformation lookup, synthesis method retrieval, compound discovery, invention recipe generation, and dataset statistics. Additional web pages include formula_synthesis_web.html (formula-to-synthesis explorer with success probability display) and how_to_transformation_web.html (transformation browser).

The BluRayDataReader class handles gzip decompression, caching (configurable cache size, default 1,000 entries), and index lookups transparently. For deeper analysis, import the core engine directly: from alchemy_data_v2 import AlchemyDataV2.

Run USE_DATA.bat or python bluray_data_reader.py interactive to enter a menu-driven command-line session. The menu offers compound search, transformation lookup, synthesis methods, compound discovery, recipe generation, and statistics display — the same capabilities as the web interface, without a browser.

RUN_BLURAY_GENERATOR.bat launches large-scale data generation. CHECK_PROGRESS.bat monitors the run. CREATE_BURNABLE_DISCS.ps1 (PowerShell) prepares the generated data for Blu-ray burning. These tools target 25–50 GB datasets for archival distribution.

This package ships on a single Blu-ray disc, posted to the delivery address on your order. The archive on the disc is AES-256 encrypted; the passphrase is sent separately, by email, once the disc is despatched — so the disc alone is of no use to anyone who intercepts it.

Discs are despatched by tracked, signed-for post. Allow 5 business days for mastering and verification before despatch.