Previous Next Title Page Contents

VI Analysis of data coming from a frame channel: the selector

It is possible to analyze data coming from a frame channel by retrieving it sequentially or randomly. In this case, it is the user's responsibility to do the analysis loop. It may be interesting to automate the loop in order to let the user focus on the physics analysis, and in order to prepare for parallel processing.
This is done by using the frame channel VFrameChannel::Process() method which sends a restricted set of data from a frame channel to a processing script or class written by the user, called a "selector". Before we can process data, we need to make this selector.
Generation of a standard source skeleton, if "fc" is a valid frame channel, is made with the VFrameChannel::MakeSelector method :
vega[] fc->MakeSelector("selectorname");
This creates skeleton source code of a class named "selectorname" in two files selectorname.h and selectorname.C. The file selectorname.C is the only relevant for the user.
The generated class has a few initially empty methods that the user should fill :

Once the file selectorname.C is customized, it becomes the argument of VFrameChannel::Process :
vega[] fc->Process("selectorname.C", 712345678.65,1300,"");
this will interpret the file selectorname.C and send it frames starting at GPS time 712345678.65, for a duration of 1300 s. The last argument is an optional condition on the data. One can also compile the script before processing :

vega[] fc->Process("selectorname.C++");
will compile the file selectorname.C, load it and process all the data available to the frame channel through it.
When appending a single "+", the file will be compiled and loaded, but subsequent compilations will only be done if the file has changed since the last compilation.


JavaScript must be enabled to display this email address.


Last update :27/06/2003;
Previous Next Title Page Contents