skbio.stats.distance.DissimilarityMatrix¶

class skbio.stats.distance.DissimilarityMatrix(data, ids=None)[source]¶

Store dissimilarities between objects.

A DissimilarityMatrix instance stores a square, hollow, two-dimensional matrix of dissimilarities between objects. Objects could be, for example, samples or DNA sequences. A sequence of IDs accompanies the dissimilarities.

Methods are provided to load and save dissimilarity matrices from/to disk, as well as perform common operations such as extracting dissimilarities based on object ID.

Parameters:

Parameters:	data : array_like or DissimilarityMatrix Square, hollow, two-dimensional `numpy.ndarray` of dissimilarities (floats), or a structure that can be converted to a `numpy.ndarray` using `numpy.asarray` or a one-dimensional vector of dissimilarities (floats), as defined by scipy.spatial.distance.squareform. Can instead be a DissimilarityMatrix (or subclass) instance, in which case the instance’s data will be used. Data will be converted to a float `dtype` if necessary. A copy will not be made if already a `numpy.ndarray` with a float `dtype`. ids : sequence of str, optional Sequence of strings to be used as object IDs. Must match the number of rows/cols in data. If `None` (the default), IDs will be monotonically-increasing integers cast as strings, with numbering starting from zero, e.g., `('0', '1', '2', '3', ...)`.

data : array_like or DissimilarityMatrix

Square, hollow, two-dimensional numpy.ndarray of dissimilarities (floats), or a structure that can be converted to a numpy.ndarray using numpy.asarray or a one-dimensional vector of dissimilarities (floats), as defined by scipy.spatial.distance.squareform. Can instead be a DissimilarityMatrix (or subclass) instance, in which case the instance’s data will be used. Data will be converted to a float dtype if necessary. A copy will not be made if already a numpy.ndarray with a float dtype.

ids : sequence of str, optional

Sequence of strings to be used as object IDs. Must match the number of rows/cols in data. If None (the default), IDs will be monotonically-increasing integers cast as strings, with numbering starting from zero, e.g., ('0', '1', '2', '3', ...).

Notes

The dissimilarities are stored in redundant (square-form) format [R62].

The data are not checked for symmetry, nor guaranteed/assumed to be symmetric.

References

[R62]

(1, 2) http://docs.scipy.org/doc/scipy/reference/spatial.distance.html

Attributes

`T`	Transpose of the dissimilarity matrix.
`data`	Array of dissimilarities.
`dtype`	Data type of the dissimilarities.
`ids`	Tuple of object IDs.
`png`	Display heatmap in IPython Notebook as PNG.
`shape`	Two-element tuple containing the dissimilarity matrix dimensions.
`size`	Total number of elements in the dissimilarity matrix.
`svg`	Display heatmap in IPython Notebook as SVG.

Methods

`x in dm`	Check if the specified ID is in the dissimilarity matrix.
`dm1 == dm2`	Compare this dissimilarity matrix to another for equality.
`dm[x]`	Slice into dissimilarity data by object ID or numpy indexing.
`dm1 != dm2`	Determine whether two dissimilarity matrices are not equal.
`str(dm)`	Return a string representation of the dissimilarity matrix.
`copy`()	Return a deep copy of the dissimilarity matrix.
`filter`(ids[, strict])	Filter the dissimilarity matrix by IDs.
`index`(lookup_id)	Return the index of the specified ID.
`plot`([cmap, title])	Creates a heatmap of the dissimilarity matrix
`read`(file[, format])	Create a new `DissimilarityMatrix` instance from a file.
`redundant_form`()	Return an array of dissimilarities in redundant format.
`to_data_frame`()	Create a `pandas.DataFrame` from this `DissimilarityMatrix`.
`transpose`()	Return the transpose of the dissimilarity matrix.
`write`(file[, format])	Write an instance of `DissimilarityMatrix` to a file.