[R] PosixCT subsecond accuracy

Hasan Diwan hasan.diwan at gmail.com
Fri Jan 27 19:49:08 CET 2012


Never mind... options(digits.sec) is what I needed to set... -- H

On 27 January 2012 10:44, Hasan Diwan <hasan.diwan at gmail.com> wrote:
> A sample of the data I have is:
>> head(sensor)
>      logged_on    accx   accy   accz compassx  compassy compassz
> gyrox      gyroy      gyroz
> 1 1326561428000 -0.4602 0.8346 0.0936 0.145508 -0.350586 0.259766
> 59.617390  28.521740  59.617390
> 2 1326561428050 -0.4212 1.0452 0.1326 0.219727 -0.321289 0.241211
> 88.695656  27.478260  88.695656
> 3 1326561428100 -0.2496 1.3416 0.2886 0.214844 -0.326172 0.141602
> 29.565218 -14.747826  29.565218
> 4 1326561428150 -0.3276 1.0374 0.0546 0.261719 -0.314453 0.243164
> 8.765218  16.556522   8.765218
> 5 1326561428200 -0.3588 1.0686 0.2418 0.229492 -0.307617 0.212891
> -33.808697  35.756523 -33.808697
> 6 1326561428250 -0.3120 1.0686 0.2652 0.261719 -0.289062 0.246094
> -40.347828  24.626087 -40.347828
>
> The first column is epoch time. I read that POSIXct has support for
> this, sort of[1]. So, I tried:
>> sensor[,1] <- as.POSIXct(sensor[,1], origin='1970-01-01 12:00:00.000')
>> head(sensor)
>            logged_on    accx   accy   accz compassx  compassy
> compassz      gyrox      gyroy      gyroz
> 1 1970-01-01 12:00:20 -0.4602 0.8346 0.0936 0.145508 -0.350586
> 0.259766  59.617390  28.521740  59.617390
> 2 1970-01-01 12:00:10 -0.4212 1.0452 0.1326 0.219727 -0.321289
> 0.241211  88.695656  27.478260  88.695656
> 3 1970-01-01 12:00:00 -0.2496 1.3416 0.2886 0.214844 -0.326172
> 0.141602  29.565218 -14.747826  29.565218
> 4 1970-01-01 12:00:50 -0.3276 1.0374 0.0546 0.261719 -0.314453
> 0.243164   8.765218  16.556522   8.765218
> 5 1970-01-01 12:00:40 -0.3588 1.0686 0.2418 0.229492 -0.307617
> 0.212891 -33.808697  35.756523 -33.808697
> 6 1970-01-01 12:00:30 -0.3120 1.0686 0.2652 0.261719 -0.289062
> 0.246094 -40.347828  24.626087 -40.347828
>
> That reading is quite far off, a more accurate reading should be:
>> sensor[,1] <- as.POSIXct(sensor[,1]/1000, origin='1970-01-01 12:00:00')
>> head(sensor)
>            logged_on    accx   accy   accz compassx  compassy
> compassz      gyrox      gyroy      gyroz
> 1 2012-01-15 05:17:08 -0.4602 0.8346 0.0936 0.145508 -0.350586
> 0.259766  59.617390  28.521740  59.617390
> 2 2012-01-15 05:17:08 -0.4212 1.0452 0.1326 0.219727 -0.321289
> 0.241211  88.695656  27.478260  88.695656
> 3 2012-01-15 05:17:08 -0.2496 1.3416 0.2886 0.214844 -0.326172
> 0.141602  29.565218 -14.747826  29.565218
> 4 2012-01-15 05:17:08 -0.3276 1.0374 0.0546 0.261719 -0.314453
> 0.243164   8.765218  16.556522   8.765218
> 5 2012-01-15 05:17:08 -0.3588 1.0686 0.2418 0.229492 -0.307617
> 0.212891 -33.808697  35.756523 -33.808697
> 6 2012-01-15 05:17:08 -0.3120 1.0686 0.2652 0.261719 -0.289062
> 0.246094 -40.347828  24.626087 -40.347828
>
> But now the fractional seconds are missing... Help? -- H
> --
> Sent from my mobile device
> Envoyait de mon portable
>
> 1. https://stat.ethz.ch/pipermail/r-devel/2009-March/052621.html



-- 
Sent from my mobile device
Envoyait de mon portable



More information about the R-help mailing list