Author: Zachary Pincus zpincus@gmail.com

Python modules for common tasks in the Pincus lab.

Functions for computations over plane curves, approximated as series of points (polylines) or parametric splines.

• curve.geometry: basic algorithms for polyline curves.
• curve.interpolate: methods for resampling polylines and fitting them to smoothing splines (using scipy.interpolate.fitpack).

Functions for basic image processing.

• image.colorize: color-map and color-tint grayscale images to RGB color, and combine color images. Also convert wavelengths to colors.
• image.fft: construct simple Butterworth filters for filtering image frequencies by FFT.
• image.fast_fft: use pyfftw to speed up FFT image filtering.
• image.mask: basic mask-processing functions to demonstrate the uses of scipy.ndimage's binary morphology operators.
• image.polyfit: fit image intensities to a low-order polynomial as a way of estimating the image background.
• image.resample: transform an image into the frame of reference of an aribrtrary rectangle or spline (defined using curve.interpolate).
• image.sample_texture: sample image patches for use in texture-based classification

Functions for analysis of 1-dimensional data.

• scalar_stats.hmm: estimate paths through a HMM with the Viterbi algorithm and estimate HMM parameters from observed data.
• scalar_stats.kde: estimate the distribution or survival function of a set of data using scipy.stats.kde
• scalar_stats.mcd: calculate the robust mean and standard deviation of data with the univarite MCD estimator of outliers vs. inliers.
• scalar_stats.moving_mean_std: estimate smooth trendlines for mean and standard deviation from scattered data in x and y.
• scalar_stats.smoothing: calculate weighted means of data, or smooth scattered data with Savitzky-Golay, LOWESS, or robust polynomial fitting.