can matlab handle data with microsecond precision and use it to create graph?

I'm using matlab to write graphic of stock price. On X I just use "number of microsecond elapsed from certain moment". So this is just integer value so matlab is easy to handle it. But it is not convenient because every time I have to "convert" "number of elapsed microseconds" to "time". I would prefer to see actual "time" instead.

Can matlab handle something like that "16:15:13:460882"? I need to create graph from the series of "time, value".

If so what format should I use? I need Matlab to understand that this is "time"

Answer 1

[Walter Roberson]

Unfortunately, MATLAB's date processing routines only extend to milliseconds; see http://www.mathworks.com/help/techdoc/matlab_prog/bspgcx2-1.html#bspgc4m-1

It could -- there is enough precision available to handle milliseconds up to year numbers above 100,000 -- but it does not.

You can apply your own time labels to a graph by setting XTickLabel or YTickLabel cell arrays of strings.

[Oleg]

actually I just realized that "milliseconds precision" is exactly what I actually need. Will matlab automatically recognize columns as "date" if I'm using dlmread function to read context of the file to variable and then I use plot to create graph? Thanks.

[Walter Roberson]

No, dlmread() expects numeric data. textscan() would give more flexibility. There is no "date" format field, though, so you would need to read the field as a string and then datenum() the strings.

There is a possibility that importdata() might recognize the file.