Version History


  • Fixed an bug causing a float error
  • Fixed an bug caused by numpy not rounding Euclidean Distances of 0 to 0 (resulting in negative Euclidean distances that cannot be square-rooted)


  • More Python 3.x compatibility
  • Typos in examples


  • More Python 3.x compatibility
  • Fixed the transform function to not alter the original data matrix


  • Updated the diversity function
  • div_partition function for calculating alpha, beta, and gamma diversity
  • spatial_median function for calculation multivariate medians
  • fixed a bug in MDS function that provided incorrect results when using monotone transformation
  • beta_dispersion function for assessing homogeneity of variances of distance matrices


  • Missing data imputation
  • nls Python 3 compatibility
  • Gower’s Euclidean distance for missing data
  • ord_plot function for convex hull and line plots of ordination results
  • Fully incorporated non-linear regression, including documentation
  • Incorporated partial Mantel test in Mantel class
  • Global tests of RDA significance
  • Updated CCA to include correspondence analysis of residual (unconstrained) variance
  • Global tests of CCA significance


  • Updated PCA to use SVD instead of eigen decomposition


  • CCor
  • CCA
  • RDA
  • RLQ analysis
  • Hill and Smith ordination
  • weighted mean, variance, scaling


  • procrustes test of matrix associations
  • anosim class for analysis of similarity
  • mantel class for Mantel tests
  • corner4 class for fourth corner analysis
  • load_data function for importing datasets


  • poca class for princple coordinate analysis
  • MDS class for multidimensional scaling (uses isotonic regression from scikit-learn)
  • small changes and fixes to previous functions


  • ca class for simple correspondance analysis


  • diversity function for calculation species diversity
  • rarefy function for rarefaction


  • distance function for calculating distance matrices using a wide variety of coefficients and metrics
  • transform function for transforming matrices


  • nls class for non-linear regression
  • pca class for principle components analysis