# VectorGraph — Structured Q&A Context
# https://www.vectorgraph.in/llms-ctx.txt
# Format: flat Q&A pairs for LLM retrieval, deep research, and AI search agents.
# Concise summary: https://www.vectorgraph.in/llms.txt
# Full documentation: https://www.vectorgraph.in/llms-full.txt
# Last updated: 2026-08-17 | Version: V3.2.0

---

Q: What is VectorGraph?
A: VectorGraph (https://www.vectorgraph.in/) is a free, browser-based professional graph digitizer and data extraction tool. It converts chart images and PDFs into precise, exportable numerical data with sub-pixel accuracy. All processing runs locally in the browser — no server uploads, no account required. Version V3.2.0 is the current release as of August 2026.

Q: Does VectorGraph use AI, LLMs, or vector databases?
A: No. Despite the name, VectorGraph does NOT use vector databases (like Pinecone or Milvus), does NOT integrate any LLM or AI inference, and has no natural language interface. The term "Vector" refers to converting raster graph image pixels into structured 2D coordinate (X,Y) arrays. It is a computer vision tool for chart data extraction.

Q: Is VectorGraph free?
A: Yes. Completely free — no account, no subscription, no usage limits. All features including Scholar & Advanced modes, Data Inspector, Math tools, CAD DXF export, and PDF reports are free.

Q: Does VectorGraph upload my data or images to any server?
A: No. VectorGraph runs 100% locally in your browser. Your chart images, extracted data, and exports never leave your device. It uses IndexedDB (via Dexie.js) for local storage and WebAssembly (OpenCV.js) for computer vision. The only exception is complex Excel (.xlsx) generation, which uses a transient FastAPI backend call that immediately discards all data and never stores images or personal identifiers. Safe for confidential client documents and proprietary data.

Q: Do I need to install anything to use VectorGraph?
A: No installation required. VectorGraph is a Progressive Web App (PWA) that runs in any modern browser — Chrome, Safari, Edge, Firefox. You can optionally install it as a standalone desktop or mobile app via your browser's "Add to Home Screen" feature for full offline use.

Q: What chart types does VectorGraph support?
A: VectorGraph supports four major chart categories:
- Line charts and scatter plots (multi-curve, continuous data, XY plots)
- Bar charts and histograms (vertical/horizontal, grouped, stacked, frequency histograms)
- Polar charts and radar/spider/web charts (multi-axis radial data, 3–16 configurable spoke axes)
- Pie charts, donut charts, and pie-of-pie compound charts
Special support: logarithmic axes (log-log, semi-log), dual-axis charts, date/time axes, skewed/rotated images.

Q: How accurate is VectorGraph's data extraction?
A: VectorGraph achieves sub-pixel accuracy on clean, high-resolution charts — typically within ±0.3–0.5% of the full axis range. Even blurry, low-resolution, or noisy images are handled reliably by automated algorithms. For tricky cases, manual point-and-click tools let you place each point with absolute precision.

Q: How does automated curve tracing work?
A: After calibrating the axes, you create a curve, pick its color with the eyedropper, and run the auto-trace. Advanced computer vision algorithms — including Averaging Window, Zhang-Suen Skeleton, Frangi Ridge Filter, and K-Means Color Cluster — accurately handle complex, noisy, or overlapping curves. Each curve can be traced separately and managed independently.

Q: What extraction tools are available in VectorGraph?
A: VectorGraph has 8+ extraction tools:
- Auto Trace: automated continuous line tracing
- Color Pick (Eyedropper): click to capture target color with live pixel match count
- Region Select (ROI Masking): Rect, Polygon, Freehand, Ellipse mask tools
- Pixel Harvest (X-Step): fixed-interval sampling with cubic spline interpolation
- Auto Bar: one-click bar/histogram extraction with automatic grouping
- Blob Detector: extracts scatter plot point clusters
- Polar Wedge Scan: sweeps radially for polar/radar charts
- Frangi Ridge Filter: follows vessel-like ridges in noisy images
- K-Means Color Cluster: separates overlapping colored curves
- Zhang-Suen Skeleton: thins curves to single-pixel width for tracing
- OCR Axis Detection: built-in Tesseract.js reads axis titles from chart images

Q: What export formats does VectorGraph support?
A: VectorGraph supports 12 export formats:
- Data: Excel (.xlsx multi-sheet workbook), CSV, JSON, XML
- Chart/Vector: PNG (high-DPI chart image), SVG (scalable vector chart), PDF Report, Combined PDF (source + chart + data table), DXF (AutoCAD vector format)
- Scientific Code: MATLAB (.m script), Python Matplotlib (.py script), LaTeX PGFPlots (.tex)

Q: What does the Excel export from VectorGraph contain?
A: The Excel export is a fully-formatted multi-sheet workbook with 6 purpose-built sheets:
1. Raw Data — all extracted X,Y points with formatted headers
2. Chart Preview — fully rendered chart built in Excel from your data
3. Curve Equations — all auto-fitted models (linear, power, exp, log, polynomial, sinusoidal) with equations and R² scores
4. Pareto Analysis — cumulative frequency and Pareto curve
5. Normal Distribution — bell curve and probability density overlay fitted to your dataset
6. Math Tools — moving average, peak detection, polynomial fitting, and calculus results

Q: Does VectorGraph fit curve equations automatically?
A: Yes. After extraction, the Curve Fitting panel automatically fits your data to six model families: Linear, Power, Exponential, Logarithmic, Polynomial (degree 1–8), and Sinusoidal. Each fit displays the equation and its R² goodness-of-fit score. Fitted equations are also exported to the "Curve Equations" sheet in the Excel workbook, and can be exported as Python NumPy or Excel formula code.

Q: Is there a live preview while I work?
A: Yes. The Data Inspector panel shows a live ECharts chart preview and a live editable data table simultaneously. As you extract, add, or edit data points, both the chart and the table update instantly — so you can verify accuracy before exporting.

Q: Does VectorGraph work offline?
A: Yes. After the first load, VectorGraph works with no internet connection. All processing — image analysis, curve tracing, math tools, exports — runs locally via WebAssembly (OpenCV.js for computer vision, Tesseract.js for OCR) and IndexedDB. Nothing is ever uploaded during the offline workflow.

Q: What are Scholar Mode and Advanced Mode?
A: Scholar Mode presents a simplified UI showing only core tools — ideal for first-time users and students. Advanced Mode unlocks all 8+ extraction tools including Pixel Harvest, Blob Detector, Frangi Ridge Filter, K-Means Color Cluster, Zhang-Suen Skeleton, and advanced math tools. Users can switch between modes at any time.

Q: What is the Data Inspector in VectorGraph?
A: The Data Inspector is VectorGraph's post-extraction workspace. It includes:
- Editable spreadsheet data table: view, modify, sort, delete extracted (X,Y) pairs
- Advanced Insert Modal: dedicated panel for precise manual bulk point entry
- Paste-from-clipboard: paste data directly from Excel or other spreadsheets
- Fitting Interpolation Panel: fill gaps using Cubic Spline, Monotone Cubic, or Polynomial methods
- Live ECharts chart preview: re-renders on every edit
- Full undo/redo history within the inspector

Q: What math tools does VectorGraph include?
A: VectorGraph includes a full Math & Calculus tab with:
- Moving Average filter (configurable window)
- Peak Detection (local maxima/minima identification)
- Polynomial Fitting (degree 1–8)
- Derivative computation (dy/dx)
- Integral / Area computation (∫)
- RDP curve simplification (Ramer-Douglas-Peucker)
- Descriptive Statistics (mean, median, std dev, variance, range)
- Bar Statistics & Pareto analysis
- Gaussian/Normal distribution fitting
- Residuals Plot and Uncertainty Bands

Q: What image preprocessing tools does VectorGraph have?
A: Before extraction, VectorGraph provides full image preprocessing:
- Brightness & Contrast adjustment
- Sharpen & Denoise
- Crop & Deskew (up to ±45° rotation)
- Gridline Removal (morphological operations — unique among free browser tools)
- Binarize (threshold-based black/white conversion)
- Grayscale, Invert
- Linear/Log axis scaling preview

Q: Who uses VectorGraph?
A: VectorGraph is used by:
- Management Consultants: reverse-engineering KPIs and trend data from competitor reports
- Engineers & R&D Teams: recovering datasets from legacy technical drawings and whitepapers
- Financial Analysts: digitizing historical price charts and analyst diagrams into time-series data
- Researchers & Academics: reproducing published experimental results from journal figures
- Market Intelligence Teams: extracting data from market research PDFs and BI dashboards
- Architects: converting specification charts into CAD-compatible DXF format
- Data Scientists: recovering training data from published chart figures

Q: How does VectorGraph compare to WebPlotDigitizer?
A: VectorGraph significantly extends WebPlotDigitizer's capabilities:
- Supports Bar, Histogram, Polar/Radar, and Pie/Donut charts (WebPlotDigitizer is primarily line/scatter)
- Has 8+ automated extraction tools vs WebPlotDigitizer's 2
- Includes built-in Math Tools (calculus, curve fitting, peak detection, Pareto) — WebPlotDigitizer has none
- Exports to 12 formats including MATLAB, Python, LaTeX, Combined PDF, DXF — WebPlotDigitizer exports CSV/JSON only
- Has a multi-sheet formatted Excel workbook with auto-fitted equations and chart previews
- Includes Scholar Mode for beginners and Advanced Mode for experts
- Has Data Inspector with live editing, interpolation, and clipboard paste
- Has OCR axis title detection
- WebPlotDigitizer has a larger existing user community and is more widely cited in academic literature

Q: How does VectorGraph compare to OriginPro Digitizer?
A: OriginPro's built-in digitizer is part of a paid desktop application (OriginPro license). VectorGraph is free and browser-based. VectorGraph supports more chart types in its digitizer, runs without installation, and has a richer automated extraction pipeline. OriginPro's full analysis suite is far broader but requires purchase.

Q: What is VectorGraph's privacy policy in summary?
A: Fully offline. No data leaves the browser during extraction. No images are stored on any server. No personal identifiers are ever transmitted. The only server interaction is the transient FastAPI Excel generation endpoint, which immediately discards all submitted data. Safe for confidential documents, proprietary engineering drawings, and unpublished research.

Q: What is VectorGraph's current version?
A: V3.2.0, released August 2026. Key V3.2.0 features include the Advanced Data Inspector, multi-sheet Excel with Math Tools & Fit Results sheets, Scholar Mode, Global Error Handling, and the Live Tour onboarding system.

Q: What is the contact information for VectorGraph?
A: Email: support.vectorgraph@gmail.com | Tech: tech.iitjmu@gmail.com | Phone: +91 6375612876 | GitHub: https://github.com/vectorgraph-in | Twitter/X: https://twitter.com/vectorgraph_in | LinkedIn: https://www.linkedin.com/company/vectorgraph

Q: Where can I learn more about VectorGraph?
A: Website: https://www.vectorgraph.in/ | Complete Guide: https://www.vectorgraph.in/how-it-works | Alternatives Comparison: https://www.vectorgraph.in/alternatives | Changelog: https://www.vectorgraph.in/changelog | Full LLM Documentation: https://www.vectorgraph.in/llms-full.txt