praxis

name	praxis
description	Scientific data analysis and publication-quality plotting. Natural-language interface for characterisation techniques — from raw data to journal-ready figures.

Praxis — Scientific Data Analysis & Plotting

Praxis (from Greek praxis — practice, action) is a scientific data analysis skill that transforms raw lab data into publication-quality figures and analysis results.

Commands

Plotting & Visualisation

• /praxis:plot — Create any plot from data. Auto-detects best plot type, or specify explicitly.
  • Example: /praxis:plot my_data.csv as scatter with error bars
  • Example: /praxis:plot xrd_scan.xy line plot, Cu Ka, 2theta range 10-80
• /praxis:compare — Overlay multiple datasets for comparison.
  • Example: /praxis:compare sample_A.csv sample_B.csv sample_C.csv as line plot
• /praxis:style — Set journal style for all subsequent plots.
  • Example: /praxis:style nature or /praxis:style acs
• /praxis:export — Export figures at publication specifications.
  • Example: /praxis:export last_figure as svg 300dpi
• /praxis:multipanel — Create multi-panel figures with (a), (b), (c) labels.
  • Example: /praxis:multipanel 2x2 grid from these four datasets

Data Processing

• /praxis:fit — Curve fitting. Auto-suggests model or specify.
  • Example: /praxis:fit gaussian to peak at 28 deg
  • Example: /praxis:fit linear to 200-400 C region
  • Example: /praxis:fit "a * exp(-b * x) + c" custom equation
• /praxis:peaks — Peak detection, fitting, deconvolution.
  • Example: /praxis:peaks find all peaks above 5% intensity
• /praxis:baseline — Baseline correction and subtraction.
  • Example: /praxis:baseline polynomial order 3
  • Example: /praxis:baseline ALS
• /praxis:fft — FFT, power spectrum, filtering.
  • Example: /praxis:fft low-pass filter at 100Hz
  • Example: /praxis:fft show power spectrum of signal.csv
• /praxis:smooth — Data smoothing and noise reduction.
  • Example: /praxis:smooth savgol window 11 order 3
  • Example: /praxis:smooth gaussian sigma 3
• /praxis:stats — Statistical analysis and error propagation.
  • Example: /praxis:stats compare groups A B C with ANOVA
  • Example: /praxis:stats descriptive on column B
  • Example: /praxis:stats normality test on data.csv
• /praxis:interpolate — Interpolation, resampling, derivatives, integration.
  • Example: /praxis:interpolate cubic spline to 500 points
  • Example: /praxis:interpolate derivative of column B
• /praxis:normalise — Data normalisation.
  • Example: /praxis:normalise min-max or /praxis:normalise area
• /praxis:batch — Process multiple files with same analysis pipeline.
  • Example: /praxis:batch all .csv in ./data/ — plot and extract peak positions
• /praxis:template — Save/load analysis templates (pipeline configs).
  • Example: /praxis:template save as "xrd_standard"
  • Example: /praxis:template load "xrd_standard" on new_data.csv
• /praxis:report — Auto-generate analysis summary report.
  • Example: /praxis:report generate for all analyses in this session

Technique-Specific Analysis

• /praxis:xrd — XRD analysis: phase ID, Scherrer size, Williamson-Hall, d-spacing.
• /praxis:dsc — DSC/TGA analysis: Tg, Tm, enthalpy, crystallinity, DTG, decomposition.
• /praxis:impedance — EIS: Nyquist, Bode, equivalent circuit fitting, Arrhenius.
• /praxis:mechanical — Stress-strain (modulus, yield, UTS, toughness), DMA (Tg, tan delta).
• /praxis:spectro — FTIR/Raman/UV-Vis: peak assignment, Tauc plot, Beer-Lambert, ATR correction.
• /praxis:xps — XPS: survey scan, high-res fitting, Shirley background, atomic %, BE calibration.
• /praxis:sem — SEM/EDS: grain size, porosity, elemental composition, line scans.
• /praxis:afm — AFM: roughness (Ra, Rq, Rz), height profiles, 2D surface analysis.
• /praxis:piezo — Piezoelectric: d33, P-E loops, S-E butterfly, resonance (kp, kt, Qm).
• /praxis:dielectric — Permittivity, loss tangent, Cole-Cole, Curie-Weiss, AC conductivity.
• /praxis:thermal — Thermal conductivity: laser flash, steady-state, k vs T.

Help

• /praxis:help — Show all commands with examples.

How It Works

1. Load data — auto-detects file format (CSV, Excel, TXT, .xy, JSON, JCAMP-DX, HDF5, .mat), headers, units, delimiters.
2. Analyse — technique-aware processing with sensible defaults.
3. Plot — publication-ready figures using journal-specific styles. Supports line, scatter, bar, errorbar, histogram, box, violin, contour, heatmap, polar, waterfall, ternary, Smith chart, broken axis, inset, multi-panel.
4. Export — PNG (300/600 dpi), SVG, PDF, EPS, TIFF with metadata sidecar.
5. Report — auto-generate markdown summary of analyses performed.

Design Principles

• Natural language first: describe what you want in plain English.
• Auto-detection: infers data type, columns, plot type, and analysis method.
• Publication-ready by default: correct fonts, axis labels, units, colourblind-safe colours (Okabe-Ito).
• Technique-aware: knows what analysis makes sense for each data type.
• Non-destructive: never modifies original files. Outputs to separate directory.
• Reproducible: every figure includes metadata sidecar for exact recreation.
• Batch-friendly: same analysis across multiple files trivially.
• Templatable: save analysis pipelines, replay on new data.

Trigger Patterns

Use this skill when:

• User has scientific/lab data and wants to plot or analyse it
• User mentions characterisation techniques (XRD, DSC, TGA, EIS, FTIR, Raman, XPS, SEM, AFM, DMA, etc.)
• User wants publication-quality figures or journal-formatted plots
• User asks to fit curves, find peaks, correct baselines, or do FFT on experimental data
• User mentions journal formatting (Nature, Science, ACS, Elsevier, etc.)
• User has voltage, temperature, force, pressure, current time-series data
• User wants statistical analysis (t-test, ANOVA, regression, normality)
• User wants to batch-process multiple data files
• User says "praxis" or uses any /praxis:* command

Scripts Location

Praxis is a Python package. Install once with pip install praxis-sci, or run from a clone via pip install -e ~/Documents/Praxis. Import with:

from praxis.core.loader import load_data
from praxis.core.plotter import plot_data

Journal Styles Available

nature, science, acs, elsevier, wiley, rsc, springer, ieee, mdpi

Supported Data Formats

CSV, TSV, TXT, Excel (.xlsx/.xls), JSON, .xy, .dat, .asc, .spe, .jdx/.dx (JCAMP-DX), HDF5 (.h5/.hdf5), MATLAB (.mat), .brml (Bruker XRD), .dta (Gamry), .mpr (Bio-Logic), clipboard