Navigation Banner
 
  HyPhy Documentation: Result Processing: Variance Estimates/Save Covariance Matrix

      Summary:  Variance estimates is a result processing tool which displays the expected variances of all the independent parameters of any of the likelihood functions of the most recent analysis. Covariance Matrix saves the entire covariance matrix to a file.

     Options: If more than one likelihood function is available, a prompt will appear, listing all the likelihood functions.
    There is also a choice as to which method to use in computing the Hessian matrix:

  • Crude A faster method which uses 1st order mixed partial derivatives approximations.
  • Finer A slower method which uses 2nd order mixed partial derivatives approximations (roughly 4 times slower than Crude).

     Output:  For Variance Estimates a table of expected parameter variances is printed to the console along with the 95% confidence interval.

         Example:

	        VARIANCE ESTIMATES

	+-----------------------+---------+------------+--------------------+
	| Parameter             |  Value  |  Std. Dev. | 95% Conf. Interval |
	+-----------------------+---------+------------+--------------------+
	| alpha                 |   0.1957|  0.02472982|(  0.1472,  0.2441) |
	+-----------------------+---------+------------+--------------------+
	| R                     |  26.4535|  0.14911572|( 26.1612, 26.7457) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Human.a     |   0.0089|  0.00251987|(  0.0040,  0.0139) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Chimpanzee.a|   0.0112|  0.00274794|(  0.0058,  0.0166) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Gorilla.a   |   0.0114|  0.00372113|(  0.0041,  0.0187) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Orangutan.a |   0.0525|  0.01182611|(  0.0293,  0.0757) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Gibbon.a    |   0.0809|  0.01566626|(  0.0502,  0.1116) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Node5.a     |   0.0197|  0.00783575|(  0.0044,  0.0351) |
	+-----------------------+---------+------------+--------------------+
	| givenTree.Node3.a     |   0.0054|  0.00304295|(  0.0000,  0.0114) |
	+-----------------------+---------+------------+--------------------+

      For Covariance Matrix, the full covariance matrix is saved to a file.

         Example:

        Order of parameters:
K tr.1.a tr.2.a tr.Node1.a tr.3.a tr.4.a
{{ 0.0082979996 ,-0.00045364725,-7.3138357e-05,-0.00016120148,-0.00037530438,-0.00014334637}
{-0.00045364725, 0.00049541356,-9.7033462e-05, 5.9047066e-05, 3.8177046e-05,-1.8211741e-05}
{-7.3138357e-05,-9.7033462e-05, 0.00020498417,-6.1730065e-05,-1.0001024e-05, 1.3160608e-05}
{-0.00016120148, 5.9047066e-05,-6.1730065e-05, 0.00048135161,-9.0633357e-05,-3.7905952e-06}
{-0.00037530438, 3.8177046e-05,-1.0001024e-05,-9.0633357e-05, 0.00049922709,-7.0732617e-05}
{-0.00014334637,-1.8211741e-05, 1.3160608e-05,-3.7905952e-06,-7.0732617e-05, 0.00062770996}}

This matrix can be imported into a spreadsheet as a comma separated file.

 
Sergei L. Kosakovsky Pond and Spencer V. Muse, 1997-2002