Frame covariance_matrix

---

covariance_matrix(self, data_column_names, matrix_name=None)

Calculate covariance matrix for two or more columns.

Parameters:

data_column_names : list

The names of the column from which to compute the matrix. Names should refer to a single column of type vector, or two or more columns of numeric scalars.

matrix_name : unicode (default=None)

The name of the new matrix.

Returns:

: <bound method AtkEntityType.__name__ of <trustedanalytics.rest.jsonschema.AtkEntityType object at 0x7f9e686f3fd0>>

A matrix with the covariance values for the columns.

This function applies only to columns containing numerical data.

Examples

Consider Frame my_frame1, which computes the covariance matrix for three numeric columns:

>>> my_frame1.inspect()

  col_0:int64    col_1:int64   col_3:float64
\--------------------------------------------\
    1            4             33.4
    2            5             43.7
    3            6             20.1

>>> cov_matrix = my_frame1.covariance_matrix(['col_0', 'col_1', 'col_2'])
>>> cov_matrix.inspect()

  col_0:float64    col_1:float64   col_3:float64
\------------------------------------------------\
     1.00             1.00            -6.65
     1.00             1.00            -6.65
     -6.65           -6.65            139.99

Consider Frame my_frame2, which computes the covariance matrix for a single vector column:

>>> my_frame2.inspect()

  State:unicode             Population_HISTOGRAM:vector
\-------------------------------------------------------\
    Louisiana               [0.0, 1.0, 0.0, 0.0]
    Georgia                 [0.0, 1.0, 0.0, 0.0]
    Texas                   [0.0, 0.54, 0.46, 0.0]
    Florida                 [0.0, 0.83, 0.17, 0.0]

>>> cov_matrix = my_frame2.covariance_matrix(['Population_HISTOGRAM'])
>>> cov_matrix.inspect()

  Population_HISTOGRAM:vector
\-------------------------------------\
  [0,  0.00000000,  0.00000000,    0]
  [0,  0.04709167, -0.04709167,    0]
  [0, -0.04709167,  0.04709167,    0]
  [0,  0.00000000,  0.00000000,    0]