lp:rminc
- Get this branch:
- bzr branch lp:rminc
Branch merges
Related bugs
Bug #188776: No documentation for mincSum and friends | Undecided | Fix Committed | |
Bug #192228: mincFDR does not work with imported volumes | Undecided | Fix Committed |
Related blueprints
Branch information
Recent revisions
- 89. By mcvaneede
-
fixed a bug in mincFDR.
mincMultiDim. When calculating the thresholds at the different
significance levels, the if-statement only returned a threshold when there were at
least 2 results, instead of at least 1. - 88. By mcvaneede
-
* in the function anatLm, added the attributes "stat-type" and "df" to the output. The output can be used as an argument to anatFDR which requires these attributes.
- 87. By EJ Nikelski
-
* added a new 'labels' argument to volume.
explodeLabelVol ume to permit
specification of a subset of labels that one wants exploded out as
masks.
* fixed a volume type determination error in mincIO.writeVolumeX.
Label volumes were not being identified as such due to
issues relating to use of floating point values to represent
integers. - 86. By EJ Nikelski
-
(1) added a couple of civet convenience functions to
compute tissue volumes.
(2) added a civet convenience function to compute the
global native-to-icbm_ model rescaling factor. This
is super useful in creating volume estimates in native
space, given icbm values as input.
(3) fixed some documentation-related problems that came
out in R CMD check - 85. By EJ Nikelski
-
(1) added a bunch of civet convenience functions. All functions
are prefixed with civet.*(2) added renamed versions of both of Jason's civet functions,
matching the new naming convention. The previous versions
still exist, but have had deprecation warnings added.(3) added a new volume.* prefix for functions that do
non-statistical processing on volumes.(4) added function volume.
explodeLabelVol ume() to split, for
example, a Civet classify volume into its 4 component
volumes. Also added function volume.combineMaskVolu mes()
to, ummmm, combine any number of mask volume objects into
a unified mask volume object.(5) finally got around to figuring out how to overload the
basic S4 arithmetic operators, allowing me to now do
"+-*/" operations on MincVolumeIO objects without
having the resulting objects converted to array objects.(6) added a new "label" volume type which is auto-detected and
displayed using a spectral-ish colormap (like the functional
volumes).(7) added a new "mask" volume type which is auto-detected and
displayed using a gray-scale (really just black and white)
colormap. - 84. By EJ Nikelski
-
Changes:
(1) Added "setter" methods for mincIO class S4 objects.
The method naming follows the "getters", thus we now
have:
1. getProperty()
2. setProperty()With the addition of these methods, users can now
write code without needing to be aware of the underlying
implementation. The previous state of affairs was quite
an embarrassment, and now I can walk proudly without
shame ... or significantly less shame.
(2) Added Rd entry for setters
(3) Modified the mincIO Sweave doc to introduce and show
use of the setters. - 83. By EJ Nikelski
-
Changes:
(1) created a new set of accessor methods for all mincIO objects. This permits the
users to get access to all mincIO slot values without needing to fiddle with
or beware of the internal S4 object organization. Yes, it is much better
this way. BTW, the accessor is called mincIO.getProperty( ). (2) updated the mincIO Sweave doc to use the new mincIO.
getProperty( )
accessor method. - 82. By EJ Nikelski
-
A few assorted changes:
(1) moved all of the S4 generics into a new AllGenerics.R file
(2) incremented the sub-version number by .0.1, since we've actually made
some fairly extensive changes over the last few commits. - 81. By EJ Nikelski
-
These are mostly just fixes to clean up our "R CMD check" output a liitle.
Here's a list of the bits and pieces:(1) Added the "PACKAGE=" argument to all .C/.Call calls.
This mostly required specification of the "RMINC" package, however,
in some of the rmincLmeLoop.R functions, we needed to specify "nlme".
Note to self: How on earth is the nlme library loaded? Not loaded with
RMINC, and no explicit load command that I could see. Does this stuff
even work?(2) Added an "\VignetteIndex
Entry" line to all Sweave docs (3) Added all undocumented functions to the RMINC-internal.Rd file.
If anyone ever gets bored ... or if we want to release this package for general
usage, most of the functions contained here will need a doc page be written.(4) Fixed some mincIO S4 signatures and related documentation.
(5) Substituted all logical "T/F" with "TRUE/FALSE". While the T/F form works, it's neither
recommended not supported. CMD check winges about it.(6) Loads of fixes to classical RMINC functions and S3 methods.
- 80. By EJ Nikelski
-
Added a prefix to all mincIO methods and functions. In particular, all functions
dealing directly with mincIO objects now sport a mincIO.* prefix, common functions
that are not mincIO-specific, now have a rminc.* prefix. Why the prefix? Well,(1) Jason seemed to be moving in this direction anyway
(2) it makes what is slowly becoming a very large package somewhat more manageable
(3) I found that using the prefix makes it easier to use Tab-completion when writing
code in an editor that supports this ... as well as the R command line. It's
quite lovely to be able to type "mincIO.<tab>" and see a list of all mincIO
methods/functions. Yes, I know that Jason was trying to move away from using
the "." as a delimiter within function names due to the goofy way that S3 method
scheduling works ... but
(a) none of these functions are S3
(b) this way looks so purdy.
Branch metadata
- Branch format:
- Branch format 6
- Repository format:
- Bazaar-NG Knit Repository Format 1