每毫秒绘制时间序列

Question

I can't get my plot to show the points per millisecond. It always aggregates it all together per second. But I have data per millisecond. What am I doing wrong?

I have a data frame with:

                   time     x      y    z
2015-09-24 14:21:40.130  0.11  -0.19 1.54
2015-09-24 14:21:40.229  0.11  -0.19 1.54
2015-09-24 14:21:40.279  1.27  -0.55 1.69
2015-09-24 14:21:40.309  0.19   0.05 0.44
2015-09-24 14:21:40.342  1.15  -0.16 1.39
2015-09-24 14:21:40.426  0.03   0.88 0.12

The time column is POSIXct time

options(digits.secs=3)
data$time <- as.POSIXct(data$time, "%H:%M:%OS" ,tz = "GMT")

When I plot it plot(data$time, data$z) it takes all the points in one second and plots that on the same line. So if I plot only 4 seconds of the data frame it shows 4 stripes of points instead of a continuous distribution of the milliseconds it contains between the seconds.

I would like to see the points distributed in milliseconds. I also tried ggplot but there it is the same. Hopefully my question is a bit clear, for I can't post any image.

for image see: https://github.com/NieneB/rplot/blob/master/Rplot01.png

Answer 1

Part of the problem is scale. Millisecond differences aren't well separated when your scale covers minutes of data. I've expanded your sample to include four data sets separated by several minutes and where each set lies within the same second as is suggested by the chart you reference in your link. ggplot could be used to plot this as shown in the following:

library(ggplot2)
sp <- ggplot(cbind(data, DayHrMin=format(data$time,"%Y-%m-%d %H:%M:%S")), 
             aes(x=as.POSIXlt(time)$sec, y=z))
sp <- sp + geom_point()
sp <- sp + facet_wrap( ~ DayHrMin, scale="free_x")
plot(sp)

This generates the following chart: