occasional track errors (linc-cox2 adult, embryonic, and haunt adult)

[agroff11]

Linc-Cox2 adult and embryonic, and haunt embryonic have issues with gene tracking. I though the cummerbund method would fix this but it doesnt.

In .md output, error is: ## Error: dims [product 3] do not match the length of object [4]

And when I work back through I find that everything works (with the cummerbund makeGeneTrack method) until plotTracks, where I get "Error in n - 1 : non-numeric argument to binary operator"

traceback()
12: movingAverage(x, window) at #3
11: trans(vals)
10: .local(x, ...)
9: score(GdObject)
8: score(GdObject)
7: .local(GdObject, ...)
6: collapseTrack(GdObject, diff = diff, xrange = c(minBase, maxBase))
5: collapseTrack(GdObject, diff = diff, xrange = c(minBase, maxBase))
4: .local(GdObject, ...)
3: drawGD(trackList[[i]], minBase = ranges["from"], maxBase = ranges["to"],
prepare = TRUE, subset = FALSE)
2: drawGD(trackList[[i]], minBase = ranges["from"], maxBase = ranges["to"],
prepare = TRUE, subset = FALSE)
1: plotTracks(myTracks, from = from, to = to, chromosome = myChr,
showAxis = FALSE, background.title = "black", col.title = "white",
col.axis = "black", sizes = trackSizes, geneSymbol = TRUE)

I'm not entirely sure what's going on with these 3. Haunt adult works fine (and an iteration or two ago, so did cox2 embryonic)

thoughts?!

[Loyale]

Looks like the problem is with my movingAverage() function. Can you adjust the window size or don't do a transform for those particular genes?