DiscussWorldIssues - Socio-Economic Religion and Political Uncensored Debate - View Single Post

**JohnImamadviser** · 01-21-2008, 08:36 PM

[donald is speaking here with a follow-up from this post found in the human mind, body and cosciousness thread in new sciences: http://divinecosmos.com/forums/showt...1721#post21721 ]

i haven't checked on the measurement for a long time and was amazed that it has gone from yellow and green to blue which is 90%, wow, donald

http://noosphere.princeton.edu/gcindex.html

brad anderson wrote a java application to run on his server that pulls gcp/egg data in near real time to display a summary statistic in a color index (see below). the application runs on brad's server, and is linked from the shaded dot on the gcp homepage. here are some explanatory extracts from his email description of the process.

the server calculates the probability of chisquare (the squared stouffer z, or z²) on a second time frame basis (in real time) following the gcp analysis instructions. then a mac widget (mini application) pulls the statistical analysis from the server and presents the user with a colored button for the probability.

i have been wanting to do this concept for about 6 months now. you will be happy to know that each individual instance is not pulling data from your servers. the processing would take too long, and i was worried that if this got popular, that it could overload your servers.

the program runs the statistics continuously minute by minute. it then dumps a file of the second by second probability for the minute into a web directory. each client downloads this probability file for display. this way, my machine has the tough job of statistical analysis, while the gcp servers are isolated from the potential high volume that this could generate, and the clients don't have to spend any time processing.

the server runs 10 minutes behind. it stores the prior minute of statistics while it pulls the current minute. the clients are buffered with another minute to allow the data to be downloaded asynchronously while the client displays older data. this way there is no stopping or blocking when the data is grabbed. each has its own one minute buffer, putting the client 12 minutes behind the data generation.

the program creates a chisquare for the most recent hour (the sum of 3600 seconds' z² values). the probability of the chisquare is then calculated. this value, representing the average departure of an hour of data from expectation, is recalculated each second in a sliding window, adding a new second and dropping the trailing second. thus, the display changes very slowly, and gives an impression of the state of the network averaged over a substantial period.

01-21-2008, 08:36 PM	#1
JohnImamadviser Join Date Oct 2005 Posts 551 Senior Member	Collective Conscience [donald is speaking here with a follow-up from this post found in the human mind, body and cosciousness thread in new sciences: http://divinecosmos.com/forums/showt...1721#post21721 ] i haven't checked on the measurement for a long time and was amazed that it has gone from yellow and green to blue which is 90%, wow, donald http://noosphere.princeton.edu/gcindex.html brad anderson wrote a java application to run on his server that pulls gcp/egg data in near real time to display a summary statistic in a color index (see below). the application runs on brad's server, and is linked from the shaded dot on the gcp homepage. here are some explanatory extracts from his email description of the process. the server calculates the probability of chisquare (the squared stouffer z, or z²) on a second time frame basis (in real time) following the gcp analysis instructions. then a mac widget (mini application) pulls the statistical analysis from the server and presents the user with a colored button for the probability. i have been wanting to do this concept for about 6 months now. you will be happy to know that each individual instance is not pulling data from your servers. the processing would take too long, and i was worried that if this got popular, that it could overload your servers. the program runs the statistics continuously minute by minute. it then dumps a file of the second by second probability for the minute into a web directory. each client downloads this probability file for display. this way, my machine has the tough job of statistical analysis, while the gcp servers are isolated from the potential high volume that this could generate, and the clients don't have to spend any time processing. the server runs 10 minutes behind. it stores the prior minute of statistics while it pulls the current minute. the clients are buffered with another minute to allow the data to be downloaded asynchronously while the client displays older data. this way there is no stopping or blocking when the data is grabbed. each has its own one minute buffer, putting the client 12 minutes behind the data generation. the program creates a chisquare for the most recent hour (the sum of 3600 seconds' z² values). the probability of the chisquare is then calculated. this value, representing the average departure of an hour of data from expectation, is recalculated each second in a sliding window, adding a new second and dropping the trailing second. thus, the display changes very slowly, and gives an impression of the state of the network averaged over a substantial period.
	Quote