Try the live webapp

Run Mie scattering simulations in your browser—no installation required.

Launch the PyMieSimX webapp to configure optical setups, run parameter sweeps, explore individual particles, and download your results as CSV files. The hosted service provides direct access to the graphical interface; local installation is also available below.

Source-model note

The Gaussian source exposed by PyMieSimX is not a generalized Lorenz–Mie theory (GLMT) implementation. It is a convenience object for defining a Gaussian illumination through its numerical aperture and optical power in watts. The Gaussian and GaussianSet source options should therefore be interpreted as a practical source parameterization, not as a separate GLMT solver.

Installation

Install the GUI with PyMieSim and PyOptik with:

pip install PyMieSimX

The first pymiesimx launch downloads PyOptik’s complete RefractiveIndex.INFO snapshot. It can also be initialized explicitly with:

python -m PyOptik setup --no-progress

Launch the dashboard with:

pymiesimx

Use pymiesimx --help to see the available host, port, browser, and debug options.

Development

Install an editable checkout with:

pip install -e .

The GUI source lives in PyMieSimX/gui. Scientific calculations are provided by the installed PyMieSim package rather than duplicated here.

Testing and automation

Run the GUI test suite with:

pip install -e ".[testing]"
python -m pytest

GitHub Actions includes quality checks, GUI tests, PyPI publishing, Conda recipe publishing, and coverage deployment. The Conda recipe is in meta.yaml and the container entry point is defined in Dockerfile.

Python API

The computational API can be used without constructing the Dash application:

from PyMieSimX import run_experiment

result = run_experiment(
    source_type="GaussianSet",
    source_values={
        "wavelength": "650",
        "polarization": "0",
        "optical_power": "1e-3",
        "numerical_aperture": "0.2",
    },
    scatterer_type="SphereSet",
    scatterer_values={"diameter": "500", "material": "1.4", "medium": "1.0"},
    detector_type="None",
    detector_values={},
    measure="Qsca",
)

PyMieSimX.create_dash_app and PyMieSimX.OpticalSetupGUI remain available for applications that need the graphical interface.

Multiple measures

The Parameter Sweep Measures selector accepts several quantities in one run. For example, select Qsca, Qabs, and Qext to compare scattering, absorption, and extinction efficiencies. Measures with the same units share a plot; different units use separate panels with a shared X axis. CSV exports contain a column for each computed measure.

The Python API accepts either a single name or a list, for example run_experiment(..., measure=["Qsca", "Qabs", "Qext"]). Results include measures with every computed name; measure retains the first name for compatibility with existing single-measure callers. Size validation accounts for the additional result columns.

Background calculations and limits

Parameter sweeps are submitted to a bounded background worker and reported to the dashboard with visible queued, running, completed, and error messages beside Run. The completed-run card shows the number of result rows. Oversized sweeps are rejected with guidance when you click Run. Results are guarded by both an in-memory dataframe limit and a serialized Dash-payload limit, with estimated size validation before execution and actual size checks after computation.

Run the command-line launcher with --debug to enable detailed logs. Log messages use the format timestamp | level | logger | message and include experiment job identifiers, input configuration, queue transitions, dataframe memory usage, and serialized result sizes.

Server-side usage metrics

The dashboard can share a PostgreSQL metrics database with RosettaX. Configure the Render service with:

PYMIESIMX_USAGE_METRICS_BACKEND=postgres
DATABASE_URL=<the shared PostgreSQL URL>

PyMieSimX writes namespaced counters to the shared metrics_counters table: pymiesimx_home_page_visit_count, pymiesimx_experiment_run_count, and pymiesimx_single_run_count. Metrics are collected only when the PostgreSQL backend is explicitly configured and available. Local installations do not persist usage counters.

To show a small Webapp visits box in the homepage’s top card, set this environment variable on the Render service and redeploy:

SHOW_METRIC=true

The box is hidden by default and requires the PostgreSQL configuration above. It shows the total home-page visit count, including repeat visits, rather than unique visitors. If the database is unavailable, the box stays hidden.

Administration page

Detailed usage counters are shown on the hidden, token-protected administration page. Configure a server-side token with:

PYMIESIMX_ADMIN_TOKEN=<a long random value>

Then open /admin?token=<the same value>. The page is intentionally absent from the public sidebar and refreshes its counters every 30 seconds.