FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

Data Viewer Up

Scritto da Ryan Hunt on .

 

The new HUGnetView data viewer for the live data stream is up with the improved data viewer.  It is still catching up to the newest data.  

Eventually it will be at this address:  

http://data.hugnetlab.com

For now, you can go to this IP address, instead, 

http://96.126.123.132

 

The old location will still work, but will be even slower, now.

Add comment


Security code
Refresh

Comments   

 
0 #8 Ecco 2012-11-14 13:10
@RobWoudenberg: your suggestion is certainly useful, but people with Linux or MacOS probably can't get to use Excel for its pivot tables.
Quote
 
 
0 #7 RobWoudenberg 2012-11-14 13:07
(seems you need to click the graph once you opened it in Excel first to get the piviottable field list where you can select the desired parameters to be displayed)
Quote
 
 
0 #6 David Roberson 2012-11-13 22:52
I performed a closer look at the resistance curve for the active wire and now suspect that the periodic dips may well be caused by round off errors. The ratio of voltage to current is used to calculate the value.

Is it possible to use an ohm meter to directly make the measurement and thereby eliminate this error? The other alternative is to more tightly control the voltage being applied to the wire which is your plan I believe. We should remember the source of this noise for future reference as it tends to mask real resistance changes.

@Ecco: I was referring to data that was near the 17:00 time frame, not that exact time. Perhaps I should have been more precise.
Quote
 
 
0 #5 Ecco 2012-11-13 22:26
@David Roberson: I think there's a hole in the data set (or for some reason data went missing) and the system is interpolating the data points inbetween.
Quote
 
 
0 #4 David Roberson 2012-11-13 22:18
Something is strange happening around 17:00 today with the data. I see that the resistance of the active wire appears to have a large dip in value that occurs periodically. Also, much of the noise seems to have vanished.

Is this raw data or has some manipulation been done to it to make it look this way? I would guess that someone at the lab has adjusted the data to compensate for the voltage drop that was noted the last few days.

Has there been a change to the test system which cleans up the noise to a much larger extent? I will view the voltage waveform to see if it has also changed.

I am not being critical, just looking for unusual changes.
Quote
 
 
0 #3 David Roberson 2012-11-13 06:33
I request that you keep the data available that is collected at the fastest possible rate. Allow some of us to access this information even if most of the others are happy with the decimated versions.

The reason for this request is that no one can be confident that the excess heating mechanism acts slowly. I have a gut feeling that it is very quick and we will have a difficult time understanding the waveforms even at the current rate. The best comparison is to that of an aliased function.

Allow us to obtain the raw unfiltered data as is available now. We will apply the appropriate filtering functions to suit our purposes.
Quote
 
 
0 #2 Robert Greenyer 2012-11-13 02:54
@Ecco

Nicolas has written a piece of software to address this - it will average 10 seconds of data and post it - this will be implemented at some point and will allow for removal of noise from the signal and a better experience overall.
Quote
 
 
0 #1 Ecco 2012-11-13 02:29
One of the reasons of the large bandwidth usage you were experiencing is that the page tries to download every data point from the time range selected, even if they exceed the chart resolution.

The chart is 750 pixel wide, but at a rate of 0.5 Hz, 4 hours of data (for example) would mean that 4*60*60*0.5 = 7200 data points are being attempted to be fit there. This is almost 10 times more data than can be actually viewed. It gets even worse when one tries to display 12 hours of data. Even for only 30 minutes of data you're already exceeding the chart resolution.

In practice, it isn't even that much necessary to plot a data point every 2 seconds. One every 5 or even 10 seconds would still be enough but would save a lot of bandwidth and allow real time data streaming for everybody.

.7z compressed .csv files containing daily full data sets (or about specific events) for those interested can still be provided separately.
Quote
 

Here is your generous contributions so far towards our $500,000 target, thanks everyone! : $45,020   Please Donate
See the current state of our booked costs here