stats
Evaluate selected statistics of ARWM chain
Syntax
S = stats(pos, theta, logPost, ...)
S = stats(pos, fileName, [], ...)
Input arguments
-
pos
[ poster ]Posterior simulator object that has generated the
theta
chain. -
theta
[ numeric ]MCMC chain generated by the
poster/arwm
function. -
logPost
[ numeric ]Vector of log posterior densities generated by the
arwm
function;logPost
is not necessary if you do not requestmdd
, the marginal data density. -
fileName
[ char ]File name under which the simulated chain was saved when
poster/arwm
was run with optionssaveEvery=
' andsaveAs=
.
Output arguments
S
[ struct ]Struct with the statistics requested by the user.
Options
-
MddGrid=0.1:0.1:0.9
[ numeric ]Points between 0 and 1 over which the marginal data density estimates will be averaged, see Geweke (1999).
-
Progress=false
[true
|false
]Display progress bar in the command window.
Options to Include/Exclude Output Statistics
-
Bounds=
[true
|false
]Return the lower and upper parameter bounds set up by the user.
-
Chain=
[true
|false
]Return the entire simulated chains of parameter values.
-
Cov=
[true
|false
]Return the sample covariance matrix.
-
'
Hist=
' [ numeric | empty ]Return histogram bins and counts with the specified number of bins.
-
Hpdi=
[false
| numeric ]Return the highest probability density intervals with the specified percent coverage (e.g. 90% is entered as
90
, not0.90
). -
KsDensity=
[true
|false
| numeric ]Return the x- and y-axis points for kernel-smoothed posterior density; use a numeric value to control the number of points over which the density is computed.
Mdd=true
[ true
| false
]
Return minus the log marginal data density.
Mean=true
[true
|false
]Return the sample averages.
Median=false
[ true
| false
]
Return the sample medians.
Mode=false
[ true
| false
]
Return the sample modes based on histograms.
-
Prctile=[]
[ numeric | empty ]Return the specified percentiles.
-
Std=true
[true
|false
]Return the sample std deviations.