Undocumented behavior around daylight savings time?

Apologies for the intrusion. I am a lurker on list.

I have been working to convert a digitized signal from a matlab file
into R
for analysis and other applications. R.matlab is working fine, and it
is
easy to convert the matlab date-time number (days since year 0) into R
date-time numbers (seconds since 1970-01-01).

Unfortunately, when I cast the R date-time number into POSIXct format
it
seems to adjust silently by one hour to reflect daylight savings time,
but
I have been unable to suppress that behavior. (The problem is that the
matlab date is already in MDT, and I don't want to have to write my own
code to suppress the added hour only when local DST rules apply.)

To use the example of the R date-time number 1340717324, this is the
behavior I observe:

as.POSIXct(1340717324,origin='1970-01-01')

[1] "2012-06-26 14:28:44 MDT"

as.POSIXct(1340717324,origin='1970-01-01',tz='')

[1] "2012-06-26 14:28:44 MDT"

as.POSIXct(1340717324,origin='1970-01-01',tz='America/Denver')

[1] "2012-06-26 14:28:44 MDT"

as.POSIXct(1340717324,origin='1970-01-01',tz='MST')

[1] "2012-06-26 13:28:44 MST"

as.POSIXct(1340717324,origin='1970-01-01',tz='UTC')

[1] "2012-06-26 13:28:44 UTC"
I was ultimately able to solve the problem by casting into and out of a
character string, but that seems risky/error prone.

rdates='%Y-%m-%d %H:%M:%S'

as.POSIXct(strptime(as.character(as.POSIXct(1340717324,origin='1970-01-01',tz='UTC')),format=rdates))
[1] "2012-06-26 13:28:44 MDT"

I have read the various help entries and even investigated the
lubridates
package, but none indicate why exactly the extra hour is being added or
how
to suppress it. (Note that I tried as.POSIXlt with various settings of
isdst, none of which worked).

I'm using "R version 2.13.1 (2011-07-08)" on
"x86_64-apple-darwin9.8.0".



-- 

----------------
Samuel Brown

[[alternative HTML version deleted]]