::  Posts  ::  RSS  ::  ◂◂RSS  ::  Contact

Timestamp range in gnuplot

June 20th, 2011
gnuplot, tech  [html]
In gnuplot, you might have some data you'd like to graph that has timestamps:
     1308369601 0.15
     1308369661 0.05
     1308369721 0.43
     1308369781 0.41
     ...
  
You can see many places that to tell gnuplot your xaxis is in seconds since the epoch you use:
     set xdata time
     set timefmt "%s"
     set format x "%H:%M"     # or anything else
     set xlabel "time"
  
If you want to limit the range, you might think you could do something like:
     set xrange [1308369932:1308373241]
  
If you try this, you'll get an error like:
     "plot.gnp", line 16: all points y value undefined!
  
To fix this, you need to subtract 946684800 from the start and end of your xrange:
     set xrange [1308369932-946684800:1308373241-946684800]
  
What is going on here? It turns out that unlike everything sensible in the unix world, gnuplot calculates time in seconds since the beginning of the year 2000 instead of the year 1970. So it interprets [1308369932:1308373241] as early morning 2041-06-17 instead of early morning 2011-06-18. The magic number, 946684800, is the number of seconds between 2000-01-01 and 1970-01-01.

I understand gnuplot was trying to be platform independent, but it's pretty annoying now that the 1970 epoch has become dominant.

Recent posts on blogs I like:

Transfers from Infrequent to Frequent Vehicles

Imagine yourself taking a train somewhere, and imagine the train is big and infrequent. Let’s say it’s the commuter train from New York down the Northeast Corridor to Newark Airport, or perhaps a low-cost OuiGo TGV from Lyon to Paris. Now imagine that you…

via Pedestrian Observations January 20, 2020

Veganism and restrictive eating

I’m reading the book Intuitive Eating, which I highly recommend. I was looking for something like it that could get me back to trusting my biological hunger without worrying that I need to control myself or my weight. It’s raised my consciousness to the w…

via Holly Elmore January 17, 2020

Algorithms interviews: theory vs. practice

When I ask people at trendy big tech companies why algorithms quizzes are mandatory, the most common answer I get is something like "we have so much scale, we can't afford to have someone accidentally write an O(n^2) algorithm and bring the site d…

via Posts on Dan Luu January 5, 2020

more     (via openring)

More Posts:


  ::  Posts  ::  RSS  ::  ◂◂RSS  ::  Contact