Features
Everything you need to stop switching between chatbot, terminal, and image editors. In one place.
Use native import to parse headers and merge multiple files, or
use the assistant's file-operations mode to programmatically trim, splice, and manipulate - all the versatility of Python.
Drag in files, fix the parse, and merge them into one project - structured and reproducible, with no code to write.
Import messy files with a checked preview before anything becomes data.
Turn columns into durable variables every plot, fit, and workspace can reuse.
When you need real data surgery, describe it and the assistant writes Python code - you read and approve every line before it runs, and your raw values never leave the Mac.
Clean and reshape data with approved Python code, without uploading raw values.
Plot 2D data as X&Y scatters, lines, and histograms; plot 3D data flat as a heat map or in a true 3D scene - surface, scatter, or line. Every figure overlays as many data series and analysis results as you want.
Overlay series, fits, summaries, and dual axes in the core figure.
See distribution shape fast, with density bins and optional fits.
The same x, y, z, shown as a flat color map or in a real 3D scene.
Turn x, y, z into a readable scalar field with contours.
Read height fields in a rotatable, export-ready 3D surface.
Inspect point clouds in real 3D, with overlays and depth cues.
Trace trajectories and parametric curves through 3D space.
Show categorical composition as pies, donuts, or stacked bars.
A separate figure type, not an analysis card: the Correlations workspace screens pairwise correlations across variables. You create it with Add Series, like any other figure.
Find latent variable blocks before you build a model.
Turn strong correlations into a graph of clusters and bridges.
Analysis cards run a pre-vetted, golden-master Python implementation and write the result back as variables you can re-plot, fit again, or export.
Raw values never leave your Mac.
Run Analysis cards on an X&Y figure.
Read broad positive distributions clearly on log-log axes.
Measure exposure with baseline-subtracted trapezoidal integration over a chosen range.
Fit binned densities with Gaussian mixtures or custom formulas.
The defensible core - stated to the depth the app actually computes.
Fit tails honestly: log-binned least squares, deliberately not billed as a maximum-likelihood estimate.
Inspect tails without binning through the Python powerlaw library.
Collapse finite-size distributions and report fitted exponents with uncertainty.
Fit E-S scaling in log-log space and take the exponent.
Choose the tail cutoff with a Kolmogorov-Smirnov (KS) distance scan.
Run Analysis cards on a Heat Map or 3D figure.
Fit point-cloud density fields as Gaussian or custom surfaces.
Fit z = f(x, y) surfaces and send contours back as variables.
Assemble panels, annotate, restyle every axis and legend, and export at journal sizes and resolutions - the entire publishing step in one app, with no round-trip to an image editor.
Assemble synced panels into one exportable board.
Leave with runnable Python that regenerates the result.
Everything you would otherwise leave the app for - done where the figure lives.
Add labels, lines, boxes, and circles where the figure lives.
Restyle axes, legends, colors, fonts, and backgrounds in place.
Export journal-spec PDF, PNG, or JPG at the required size.
Why Autoplot
The usual setup is a relay: pull the run files off a server, wrangle them in a notebook, plot in another library, then patch the figure in an image editor - and run the whole chain again every time a reviewer asks for one change.
Autoplot is the native Mac workstation that does it all in one place - import and clean, the heavy-tail and scaling analysis you can defend, the journal-grade figure, and the Python that reproduces it - with your raw data never leaving your machine.
