Message-ID: <0FF99068D2A5FE47B4E705BF5A997B30A554@EX-MBX-01.worc.ac.uk>
Date: 2011-07-01T16:18:32Z
From: Martin Wilkes
Subject: quickPCNM: Computing power or build problem?
I am attempting to apply a one-dimensional PCNM using the quickPCNM function. I have a response variable ("Au"), with 2251 observations, and corresponding 'coordinates' ("time"). I know PCNM wasn't designed for time-series analysis per se but I haven't assigned the data to class "ts" so this shouldn't be any different to a one-dimensional spatial series.
> require (PCNM)
> Au.quick.PCNM <- quickPCNM (Au, time)
After 3 attempts lasting for over 24 hours each time, R seems to get hung up, displaying only the following:
'The sites form a straight line on the map.'
'Truncation level = 0.040001'
....and this is as far as it gets.
I'm running a 3GHz Pentium 4 with 2GB of RAM - not great but I would have thought adequate enough. Could it be a problem with my build?
Many thanks,
Martin Wilkes
University of Worcester