calculate_prec_from_cov#

pharmpy.modeling.calculate_prec_from_cov(cov)[source]#

Calculate precision matrix from a covariance matrix

Parameters:

cov (pd.DataFrame) – Covariance matrix

Returns:

pd.DataFrame – Precision matrix

Examples

>>> from pharmpy.modeling import calculate_prec_from_cov
>>> from pharmpy.tools import load_example_modelfit_results
>>> results = load_example_modelfit_results("pheno")
>>> cov = results.covariance_matrix
>>> cov
               POP_CL        POP_VC   COVAPGR    IIV_CL        IIV_VC         SIGMA
POP_CL   4.408600e-08  4.761930e-08 -0.000002 -0.000001  1.552150e-07  8.042430e-08
POP_VC   4.761930e-08  7.233190e-04 -0.000804  0.000050  7.174490e-05  1.467290e-05
COVAPGR -1.679560e-06 -8.041750e-04  0.007015 -0.000108 -3.935790e-05  2.922260e-05
IIV_CL  -1.090290e-06  4.989580e-05 -0.000108  0.000180 -1.863210e-05  5.049910e-06
IIV_VC   1.552150e-07  7.174490e-05 -0.000039 -0.000019  5.588920e-05 -4.497590e-07
SIGMA    8.042430e-08  1.467290e-05  0.000029  0.000005 -4.497590e-07  5.197970e-06
>>> calculate_prec_from_cov(cov)
               POP_CL        POP_VC       COVAPGR         IIV_CL        IIV_VC          SIGMA
POP_CL   2.993428e+07  22261.039122  16027.859538  203633.930854 -39113.269503 -817314.801755
POP_VC   2.226104e+04   2129.852212    260.115313    -373.896066  -2799.330946   -7697.921603
COVAPGR  1.602786e+04    260.115313    187.053488     177.987340   -205.261483   -2224.522815
IIV_CL   2.036339e+05   -373.896066    177.987340    7542.279597   2472.034556  -10209.416944
IIV_VC  -3.911327e+04  -2799.330946   -205.261483    2472.034556  22348.216559    9193.203130
SIGMA   -8.173148e+05  -7697.921603  -2224.522815  -10209.416944   9193.203130  249978.454601

See also

calculate_se_from_cov

Standard errors from covariance matrix

calculate_se_from_prec

Standard errors from precision matrix

calculate_corr_from_cov

Correlation matrix from covariance matrix

calculate_cov_from_prec

Covariance matrix from precision matrix

calculate_cov_from_corrse

Covariance matrix from correlation matrix and standard errors

calculate_prec_from_corrse

Precision matrix from correlation matrix and standard errors

calculate_corr_from_prec

Correlation matrix from precision matrix