Now converts gatkreports to properly typed R data types in gsa.read.gatkreport

-- use the general function type.convert from read.table to automagically convert the string data to booleans, factors, and numeric types as appropriate.  Vastly better than the previous behavior which only worked for numerics, in some cases.

public/R/src/org/broadinstitute/sting/utils/R/gsalib/R/gsa.read.gatkreport.R

@@ -4,11 +4,9 @@
     colnames(d) = tableHeader;
 
     for (i in 1:ncol(d)) {
-        v = suppressWarnings(as.numeric(d[,i]));
-
-        if (length(na.omit(as.numeric(v))) == length(d[,i])) {
-            d[,i] = v;
-        }
+        # use the general type.convert infrastructure of read.table to convert column data to R types 
+        v = type.convert(d[,i])
+        d[,i] = v;
     }
 
     usedNames = ls(envir=tableEnv, pattern=tableName);