An essential package with all the non-parametric tools you love and use the most.
Work on several datasets simultaneously, or switch your focus across multiple graphics and output files, all within a single interactive & intuitive graphical Windows environment. Identify important subsets of samples or variables. Handle large datasets (subject only to available memory). Merge or split data on specified criteria. Define group structures for tests and displays. Keep track of your work in a guided workflow, reproduce analysis pathways, create custom notes, and share ideas easily using the Explorer tree navigation pane in the PRIMER workspace.
Unleash built-in wizards or a Shade plot tool to create matrix displays and heatmaps, enabling direct visualisation of data values. Choose from a plethora of options to re-order samples and/or variables independently. You can maximise a pattern of seriation along each axis of the plot, constrain by groups, an auxiliary variable or a cluster dendrogram. You can apply overall transformations and/or reduce the species set to retain only a chosen number of variables, based on a variety of criteria. Customise colours, symbols and labeling options for 2-d or 3-d displays.
Perform hierarchical agglomerative cluster analysis with a choice of linkage (single, complete, group average or flexible beta). Generate dendrograms, cophenetic distances, and identify clusters with significant remaining structure via SIMPROF tests performed at every node; create a factor based on the SIMPROF results. Divisive clustering options include unconstrained (UNCTREE) or constrained (LINKTREE) methods. Achieve non-parametric k-means clustering (kRCLUSTER) for a specified number of groups, or allow groups to be chosen via an automated SIMPROF cut-off criterion.
Unleash built-in wizards or a Shade plot tool to create matrix displays and heatmaps, enabling direct visualisation of data values. Choose from a plethora of options to re-order samples and/or variables independently. You can maximise a pattern of seriation along each axis of the plot, constrain by groups, an auxiliary variable or a cluster dendrogram. You can apply overall transformations and/or reduce the species set to retain only a chosen number of variables, based on a variety of criteria. Customise colours, symbols and labeling options for 2-d or 3-d displays.
Create a configuration of inter-point distances in chosen dimension to match original dissimilarities (or their ranks), using an iterative multi-dimensional scaling algorithm from multiple initial random configurations to identify a global lowest-stress solution. PRIMER 8 offers you metric (mMDS), threshold metric (tmMDS) or non-metric (nMDS) flavours, with copious options to customise labels and symbols, overlay images, clusters, bubbles, dendrograms, trajectories, vectors, or, and more. Use bootstrapping to show confidence regions for averaged data on MDS plots. Spin, flip, expand, rotate, sequence, animate, save and share your 2-d and 3-d ordinations.
PRIMER software is the home of Analysis of Similarities (ANOSIM), a non-parametric test for differences among a priori groups of multivariate samples (levels of a factor) on the basis of rank-order dissimilarities, using the R statistic. P-values are all achieved using rigorous permutation methods, including appropriate constraints or choice of permutable units for any given factor within the context of the full design. You can test ordered or unordered factors in any design involving up to three factors having crossed and/or nested relationships. No other software does this.
Use RELATE to calculate matrix correlations and test the null hypothesis of no association between two resemblance/distance matrices (e.g., biotic and environmental). Use the BEST routine to select a subset of (e.g., environmental) variables (from a ‘fitted data’ worksheet) that will generate a distance (or dissimilarity) matrix having maximum matrix correlation with a given (‘target’, e.g., biotic) resemblance matrix. Bespoke permutation tests automatically account for the search through subsets. Create model matrices for patterns of seriation or cyclicity for RELATE tests, which can also be done within levels of another auxiliary factor.
Calculate a suite of diversity indices for species-by-sample data, including richness (S), total individuals (N), Margalef’s species richness, Pielou’s evenness (J’), Brillouin (H), Fisher’s alpha, Shannon (H’), Simpson’s evenness/dominance, Hill’s diversity numbers, and rarefaction. Build SAD curves or dominance plots for abundance or biomass. Incorporate taxonomic, phylogenetic or functional relationships among taxa to calculate average distinctness (AvTD, Delta+, Phi+), total distinctness (TTD, PD), or variation in distinctness (VarTD, Lambda+). Use randomizations to yield funnel or ellipse plots to assess representativeness from regional or master lists.
Run any one of a whole new suite of non-parametric univariate statistical tests: Wilcoxon signed-rank test, Mann-Whitney U test, Kruskal-Wallis test, Kolmogorov-Smirnov test, or bivariate tests of association. PRIMER’s implementation of these tests is novel in that all rely on robust permutation algorithms for inference and they automatically output relevant associated graphics as well.
