MaterialScope is an open-source Python workbench for reproducible materials-characterization workflows. It brings DSC, TGA, DTA, FTIR, Raman, and XRD data into one place for import, processing, comparison, visualization, and report-ready export.
- Imports common laboratory files: CSV, TXT, TSV, XLSX, and XLS.
- Helps review column mapping and data-quality warnings before analysis.
- Provides modality-aware workflows for thermal, spectral, and diffraction data.
- Lets you compare runs and export figures, data, and reports.
- Import one or more measurement files.
- Review the detected format, columns, and metadata.
- Select the relevant analysis workflow and inspect the interactive results.
- Compare runs when needed, then export the data or a report.
MaterialScope is designed to keep source data, processing choices, visualizations, and exported results connected in a single project workflow.
Prerequisites: Python 3.11+ and pip.
git clone https://github.com/utkuvibing/MaterialScope.git
cd MaterialScope
python -m venv .venvActivate the virtual environment:
# Windows PowerShell
.\.venv\Scripts\Activate.ps1
# macOS / Linux
source .venv/bin/activateInstall the project and start the app:
pip install -e .
python -m dash_app.serverFor development and testing, also install the tooling extra:
pip install -e ".[dev]"Open http://127.0.0.1:8050 in your browser.
This command starts one process serving both Dash and the FastAPI API on the same port. You do not need to start a second backend for Dash.
For a sample-data walkthrough and feedback instructions, see the early-tester guide.
The local launch runs on your computer. Workspaces are held in memory: before stopping or restarting the server, download a .scopezip project archive from the Project page and keep it on disk. Saving an analysis result within the workspace is not a durable backup. Reopen the archive from Project to resume later; keep your original measurement files separately.
Local-first does not mean every feature is offline. Configured reference-library downloads, cloud library lookups, and literature services can make network requests. See reference-library configuration for those workflows.
API authentication is optional. To enable it for protected API routes, start the combined server with your own token:
python -m dash_app.server --token "YOUR_PRIVATE_TOKEN"The bundled Dash client receives that token automatically. Setting MATERIALSCOPE_API_TOKEN alone configures the client; it does not enable server authentication. Binding to a non-loopback address without a token prints a warning but still starts the server. /health remains unauthenticated even when a token is enabled. This API token is not a browser login system.
The Windows installer retains the Streamlit application and prefers port 8501. The Electron shell is a desktop experiment that launches a backend-only service; it does not package the Dash UI.
Streamlit still has library, license, kinetics, and deconvolution pages with no equivalent Dash pages; kinetics and deconvolution are preview features. Dash library features and backend endpoints do not imply page-level parity. Streamlit remains a dependency, including through shared translations. See the Streamlit parity inventory for the gaps and migration prerequisites. Retirement requires a separate migration; no removal date is set.
Current documentation uses MaterialScope. Existing ThermoAnalyzer packaging filenames, environment aliases, and other compatibility identifiers remain unchanged.
MaterialScope is an evolving research and engineering project. Its analysis outputs support scientific workflows but do not replace expert validation, particularly for qualitative spectral and XRD interpretation.
MIT — see LICENSE.
The optional commercial-license gate is separate from the API token. Its HMAC demo secret is public, so anyone with it can forge license keys. Commercial deployments must supply MATERIALSCOPE_LICENSE_SECRET externally; rotating it invalidates existing signed licenses and requires re-issuance. A client-held shared secret still cannot provide secure distributable licensing. Asymmetric signing is future work.