linexcel¶
Data lineage analysis for Excel workbooks.
Extracts every formula, groups stretched patterns (R1C1 canonicalization), builds a dependency graph (cells, ranges, defined names, VBA), decomposes composite functions with step-by-step evaluation, and optionally documents calculations via AI.
Install¶
Quick start¶
from linexcel import analyze
result = analyze("workbook.xlsx")
result.save_html("out.html")
print(result.stats)
See the Quick start guide for more.
Or without writing any Python¶
See the command line reference.
What would you like to investigate?¶
| Goal | Start here |
|---|---|
| Explore one sheet in a large report | Use the Graph sheet selector; directly connected cross-sheet nodes remain visible. |
| Explain a particular total | Target its output cell and trace its upstream calculations. |
| Find what an input feeds | Look up nodes and their dependents in a whole-workbook analysis. |
| Resolve linked workbook inputs | Pass --refs-dir ./linked; see lineage coverage. |
| Share findings without installing a viewer | Export a standalone offline HTML report. |
| Add explanations from your chosen model | Enable optional AI documentation. |
Guide¶
- Quick start — analyse a workbook, explore it, export it
- Command line — the same, from a shell, with no install
- Lineage coverage — what is in the graph, and what is not
- HTML export — the standalone offline report
- Workbook context & screenshots — what a reader sees, not only what the file computes
- Choosing an AI provider — local runtime, hosted gateway, or your own callable; none is chosen for you
- AI documentation — provable cards, token accounting, budgets
- Languages — nine, for both the prompt and the interface
- Data handling — what leaves the machine, and when