Argonne National Laboratory

Experimental Physics and
Industrial Control System

1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  <20122013  2014  2015  2016  2017  2018  2019  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  <20122013  2014  2015  2016  2017  2018  2019 
<== Date ==> <== Thread ==>

Subject: Re: Processing And Visualization of time series data.
From: Miroslav Mihaylov <mmihay2@uic.edu>
To: Matt Newville <newville@cars.uchicago.edu>
Cc: tech-talk@aps.anl.gov
Date: Mon, 25 Jun 2012 15:53:35 -0500
On Monday 25 June 2012 11:50:53 Matt Newville wrote:

> > http://131.193.191.37/~mnm/temperatures/menu.php
> 
> That looks nice.  I like the interactive zooming,  but I'm not exactly
> sure I understand what I'm looking at.   For me, the plots seems to
> dance around quite a bit, sometimes showing a long history of a couple
> temperatures, sometimes showing nearly nothing...  Maybe I'm not using
> it correctly?

It  does  behave properly only  in  Chrome -Safari and Opera. 
The first row of buttons shows the last X minutes or  hours of  data auto 
refreshing every second.
The big check button that says "Realtime data feed"  switches that 
functionality   on and off.
The experiment is over and there is no data being collected anymore.

Overall if there are no dancing graphs double click would re-zoom to the 
entire  time interval and mouse dragging region of the graph   or clicking on 
some of the tooltips would zoom on that region.

 The interface  is not completely full-proof and  light and I am 
in process of rewriting the whole front-end completely making it  robust 
and applicable for more general use.

> If you're using PyVisa to read the meter over GPIB/USB, that suggests
> these temperatures are not Epics Variables.   Is that correct?   You
> can run a Keithley meter with Epics and read the temperatures with a
> Channel Access client at fairly high speed.  Is this something you are
> doing, or have in mind to do?

My experiment at UIC is using epics soft ioc running with 2 PVs. Keithley 2000 
and animatics smart motor and tha's how I do it there.
I used the PyEpics wit the on changes callback.
Unfinished poster type presentation for that experimental setup.

http://miro.phy.uic.edu/~mnm/jquery/poster/#section11


I had no intention to do any of that visualization while at the beam-line.

 I had some code from before where reading GPIB with python and writing 
the value every second  to a file.  So no effort  while setting this part. 
At some later time I decided to try to writing remotely   into the database  
the computer at UIC  my laptop while at the beam-line. 
I realized that although not in the same subnet I still could insert data at 
rate 10Hz.
So I just  adopted the  python script that I use for dumping the PV values 
with the onchange callback for my UIC experiment to the existing setup. 
More or less cloned and stripped down the application that I already have.

So if I were to have this web application written in a bit more general way.
I could have started visualizing the historical data immediately with no effort 
on the setup part. 


> > In general there is a need for visualization of real time data in the
> > scientific community.  However setting up and maintaining the
> > infrastructure is not a trivial matter. With this kind of centralized
> > system this problem could be significantly reduced.
> > For example for an EPICS environment such as the beamlines at APS the
> > task of visualizing small number of PV s could be reduced down to
> > running a simple python script on the client machine given that pyEpics
> > installed on the client machine.   The database server has to be setup
> > only once at one location.
> > My rough estimate is that a single modern  desktop system can serve tens
> > of clients simultaneously each individually recording at a rate of 10Hz.
> 
> Yes, pushing changes in PV values into a database server can go very
> fast.   And, yes, using a database backend is a good way to centralize
> data collection.....
> 
 
> Can you give a few more details of what you have in mind?    You might look
> at https://github.com/newville/epicsarchiver (for code) and
> http://cars9.uchicago.edu/cgi-bin/pvarch/ (for example installation)

> of a 'python/epics/mysql' data archiver with a web interface.  It
> could definitely use some attention especially for faster, more
> interactive web and graphical displays of data, etc.    I'm not sure
> it's exactly what you have in mind, but it might be worth looking at.

I was using the epics archiver  at the beginning but that project is more 
oriented towards storing huge number of PVs while I have only two.
I have adopted ideas from the epics archiver. Such as the database runs.
  I was really biased towards the real-time data display and analysis and also 
had way to many  small requirements specific to my  experiment so 
at some point I decided to write something on my own.

I think I saw here discussions on  topic of archiving PV data. 
http://www.aps.anl.gov/epics/tech-talk/2011/msg01681.php

So I guess I was asking about the other projects such as the WebPOI pointed 
out by Xihui earlier today. 




> Cheers,
> 
> --Matt Newville <newville at cars.uchicago.edu> 630-252-0431

-- 
Miroslav Mihaylov
University of Illinois at Chicago 
Department of Physics 
2336 SES 
845 W. Taylor St. M/C 273 
Chicago, IL 60607 
312 355 0225

References:
Processing And Visualization of time series data. Mihaylov, Miroslav N.
Re: Processing And Visualization of time series data. Matt Newville

Navigate by Date:
Prev: Re: Processing And Visualization of time series data. Matt Newville
Next: epics.nsls2.bnl.gov back up Michael Davidsaver
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  <20122013  2014  2015  2016  2017  2018  2019 
Navigate by Thread:
Prev: Re: Processing And Visualization of time series data. Matt Newville
Next: EPICS and MEDM on ubuntu nalaka samarasinghe
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  <20122013  2014  2015  2016  2017  2018  2019 
ANJ, 18 Nov 2013 Valid HTML 4.01! · Home · News · About · Base · Modules · Extensions · Distributions · Download ·
· Search · EPICS V4 · IRMIS · Talk · Bugs · Documents · Links · Licensing ·