Appion

I've found a bug while running projection matching refinement jobs.
On the left-hand side of the appion web page, we're asked to provide three values for:
-outer mask radius
-inner alignement radius
-outer alignement radius
These values are expected to be in Angströms and are automatically converted in pixels to be written in the software input file.
It turns out that when launching Xmipp jobs, both inner and outer alignement radii are properly converted but it's not the case for the outer mask radius that ends being smaller than the expected value. This messes up reconstructions and I would expect a similar behavior for EMAN jobs as well because this is in the input parameters section that is common to all softwares (left-hand side).

I run a test case with the following command:

~bcarr/usr/bin/appion runJob.py --jobtype=xmipprecon --endIter=10 --outerMaskRadius=305 --innerAlignRadius=220 --outerAlignRadius=305 --angSampRate=4x10:2x5:2x3:2x2 --maxAngularChange=4x1000:2x10:2x8:2x6:5x4 --maxChangeOffset=1000 --search5DShift=4x5:0 --search5DStep=2 --symmetry="Icos (2 5 3) Crowther" --percentDiscard=10 --reconMethod=fourier --ARTLambda=0.15 --fourierMaxFrequencyOfInterest=0.25 --filterConstant=0.1 --DoComputeResolution --DoLowPassFilter --modelnames=initmodel0013.spi --stackname=partlist.sel --apix=2.75821183333334 --boxsize=256 --totalpart=100 --description=test --nodes=16 --ppn=4 --rpn=4 --walltime=240 --cput=240 --localhost=amibox03.scripps.edu --rundir=/ami/data00/appion/11oct28a/recon/xmipp_recon28 --remoterundir=~dveesler/appion/11oct28a/recon/xmipp_recon28 --runname=xmipp_recon28 --projectid=340 --expId=9203 --jobid=368

and when looking into the xmipp_protocol_projmatch.py file, we see that given a pixel size of 2.76 A/pixel the outer mask radius is not properly converted while the inner and outer alignement radius are (see below). I think there might be an error in the code for the conversion of the former value.

1. Radius of spherical mask
   """ This is the radius (in pixels) of the spherical mask
   """
   MaskRadius=46

.....

1. Inner radius for rotational correlation:
   """ In pixels from the image center
   """
   InnerRadius=80

1. Outer radius for rotational correlation
   """ In pixels from the image center. Use a negative number to use the entire image.
   WARNING: this radius will be use for masking before computing resoution
   """
   OuterRadius=111