-------------------------------------------------------------------
Recipes for making:
  * plots of slowcontrols, and
  * plots of histogram means, versus run number.
-------------------------------------------------------------------   

In the past I have made some history plots of various quantities from
slow controls versus run number (described in A. below), and also
plots of histogram means versus run number (described in B. below).
This short note documents how I made these plots.  Of course it is
possible that Jingliang and KO have their own ways of making these
plots, that I don't know much about.

A. Slow controls history vs run 
------------------------------- 

1) First I make text files of info from history.  To do this, I look
up start and end times for the runs I am interested in.  Also I look
up the "Event" numbers and variable names I am intersted in.  To find
variable names run the mhist program on midtwist:
  mhist -z /twist/data_onl/elog_history/ -l

Examples of extracting data that I made my plots from are in the script:
  /twist/data13/jamieson/2005-verification/runmhist.csh

2) Convert the output from mhist into something easier to plot, and
plot them, using a root macro:  
  /twist/data13/jamieson/2005-verification/slowconvertem.C

The macro expects several data files, with particular names, as
created in my runmhist.csh to exist.  The macro uses two other root
macros: one to convert the results into a simpler form, and another to
plot them.  To convert the results a root macro function called
slowconvert, which takes as arguments a file name prefix and a run
number range min_run to max_run, is used.  The results are then
plotted using calls to another root macro function called plotdata (in
plotemup.C), which takes three or four arguments: input data file name
(from a slowconvert output), a histogram name, a histogram title, and
optionally whether this is the last page of histograms to be plotted
in the file called data2003.ps. (move it to another name).

Then you're done you have a file of plots of slow controls data versus
run number.


B. Root histogram means versus run number
-----------------------------------------

1. Make a directory which contains links to all of the root histogram
files that you want to include in your plot.  eg. for set40 see the
directory: /twist/data13/jamieson/2005-verification/set40/, which has
broken links now, but should give you an idea of the format.

2. Make a list of histograms you want to extract means and make plot,
and I don't recall if this has to be in the directory with the plots
or the directory with the script?  Anyhow an example of the histogram
list is:
/twist/data13/jamieson/2004-verification/set37/hists.list

2. Run a root macro to make a datafile of means vs run number, eg in
my directory /twist/data13/jamieson/2005-verification, I would run the
following to make the data file for set 40:
root -l 
> .L rundir.C
> rundir("/twist/data13/jamieson/2005-verification/set40/")

The output is the file:
/twist/data13/jamieson/2005-verification/set40/data2003.dat

3. Convert the data2003.dat to a form that is similar to that used for
plotting slowcontrol vs run using a script such as:

/twist/data13/jamieson/2005-verification/striphists.csh

4. Plot the results by editing the rootmacro:
/twist/data13/jamieson/2005-verification/plotemup.C, which uses the
plotdata function as in the slowcontrols vs run number plotting. 

Then you're done, and have plots in a file called data2003.ps.  (move
it to another name).

Fin.
---------------------------------------------------------------
Last modified 2005-12-13, Blair Jamieson