class skbio.math.gradient.GradientANOVA(coords, prop_expl, metadata_map, trajectory_categories=None, sort_category=None, axes=3, weighted=False)[source]

Base class for the Trajectory algorithms


coords : pandas.DataFrame

The coordinates for each sample id

prop_expl : array like

The numpy 1-D array with the proportion explained by each axis in coords

metadata_map : pandas.DataFrame

The metadata map, indexed by sample ids and columns are metadata categories

trajectory_categories : list of str, optional

A list of metadata categories to use to create the trajectories. If None is passed, the trajectories for all metadata categories are computed. Default: None, compute all of them

sort_category : str, optional

The metadata category to use to sort the trajectories. Default: None

axes : int, optional

The number of axes to account while doing the trajectory specific calculations. Pass 0 to compute all of them. Default: 3

weighted : bool, optional

If true, the output is weighted by the space between samples in the sort_category column



If any category of trajectory_categories is not present in metadata_map If sort_category is not present in metadata_map If axes is not between 0 and the maximum number of axes available If weighted is True and no sort_category is provided If weighted is True and the values under sort_category are not numerical If coords and metadata_map does not have samples in common


get_trajectories() Compute the trajectories for each group in each category and run ANOVA over the results to test group independence.