| Deletions are marked like this. | Additions are marked like this. |
| Line 41: | Line 41: |
|
|| volume1 || description || || volume2 || description || |
||ces || Contrast Effect Size = C*bhat|| ||cespct || ces expressed as a percentage of mean baseline|| ||cesvar || variance of ces|| ||t || t-ratio|| ||sig || significance based on t-test|| ||f || f-ratio (when J>1)|| ||fsig || significance based on F-test (when J>1)|| ||minsig || best significance, boferroni corrected (when J>1)|| ||iminsig || index of best significance (when J>1)|| |
| Line 45: | Line 52: |
| description |
== Notes == The significances are stored as -log10(p). For t-tests, the signifiance is given the same sign as the t-value. For F-tests, the signifiance is not given a sign. This is the only difference with stxgrinder-sess. t-test is not saved when contrast is omnibus or zomnibus An "old" version of stxgrinder-sess is available as ["stxgrinder0-sess"]. |
Index TableOfContents
Name
stxgrinder-sess - compute contrasts based on GeneralLinearModel
Synopsis
stxgrinder-sess - ??
Arguments
Required Flagged Arguments
??
Optional Flagged Arguments
-a <analysisname> |
session-level functional analysis name |
-c <contrast> [<-c contrast>] |
contrasts |
-all |
compute all contrasts for given analysis |
-space <spacename> |
space in which to average (native, tal, sph) |
-spacedir <spacedir> |
space directory (default spacename) |
-hemi <hemisphere> |
with sph space <lh rh> |
-fdofmax <dofmax> |
max dof for F-test |
-sf <sessidfile> |
... |
-df <srchdirfile> |
... |
-s <sessid> |
... |
-d <srchdir> |
... |
-help |
|
-umask <umask> |
set unix file permission mask |
-version |
print version and exit |
Outputs
ces |
Contrast Effect Size = C*bhat |
cespct |
ces expressed as a percentage of mean baseline |
cesvar |
variance of ces |
t |
t-ratio |
sig |
significance based on t-test |
f |
f-ratio (when J>1) |
fsig |
significance based on F-test (when J>1) |
minsig |
best significance, boferroni corrected (when J>1) |
iminsig |
index of best significance (when J>1) |
Description
Notes
The significances are stored as -log10(p).
For t-tests, the signifiance is given the same sign as the t-value.
For F-tests, the signifiance is not given a sign. This is the only difference with stxgrinder-sess.
t-test is not saved when contrast is omnibus or zomnibus
An "old" version of stxgrinder-sess is available as ["stxgrinder0-sess"].
Examples
Example 1
??
Example 2
??
Bugs
The program occasionally hangs when performing the F-test. This is caused when the DOF is very large (eg, > 1000). This ofen happens when analyzing group fixed effects. If the program appears to hang, reduce FtestDOFMax with -fdofmax until it seems to run ok. You can also just skip the F-test with -noftest.
See Also
["othercommand1"], ["othercommand2"]
Links
Methods Description
description description
References
["References/Lastname###"]
Reporting Bugs
Report bugs to <analysis-bugs@nmr.mgh.harvard.edu>