This must've been brought up before, but the search didn't turn up a solution. On 3/13, 1:59:00 should be followed by 3:00:00. Intraday data, however, did not. For the heart rate, it moved from 8:53:00 to 9:53:00, a reduction of 59 min, not quite an hour. For others, all 1440 data points for 24 hours are present. (Where did the extra hour come from?). As result, R time series function, apparently expecting PDT switchover at 2AM, thinks that the time index is afu.
To see how it handled PST switch over, I went back to 2015/11/02 data. There, it only had 24 hours worth data instead of 25; 1:00:00 - 1:59:00 did not repeat as I was expecting.
Does anyone have a solution dealing with this daylight savings mess?
