Hi All,
I have made the fast normal distribution random number generator. The server is named randomNoGen.exe and takes one parameter the port number. Like speedySpline if you don't provide a parameter the port defaults to 27015. The example test app is using 27016.
Basically you send a command: 1, the mean and the standard deviation and you get back the values as doubles. The random number generator is seeded with the computer clock so you will get a different set of number for each run. If you wish to use the same number set for different runs of you program then get the numbers once and save that number set to a file.
The raw results of a single run for 20,000 Values. As you can see it takes about 0.35 seconds to do 20,000 points. The second screen shot shows a histogram of that data.
As you can see the data is pretty close: mean given = 1.0 data mean = 0.999, standard deviation given = 2.5 data SD = 2.515
Sadly I can't provide the source code for the random number algorithm as it is proprietary, however I have attached the B4J test program which include randomNoGen.exe in the objects directory.
Hope you find this useful.
Best regards
Rob
I have made the fast normal distribution random number generator. The server is named randomNoGen.exe and takes one parameter the port number. Like speedySpline if you don't provide a parameter the port defaults to 27015. The example test app is using 27016.
Basically you send a command: 1, the mean and the standard deviation and you get back the values as doubles. The random number generator is seeded with the computer clock so you will get a different set of number for each run. If you wish to use the same number set for different runs of you program then get the numbers once and save that number set to a file.
The raw results of a single run for 20,000 Values. As you can see it takes about 0.35 seconds to do 20,000 points. The second screen shot shows a histogram of that data.
As you can see the data is pretty close: mean given = 1.0 data mean = 0.999, standard deviation given = 2.5 data SD = 2.515
Sadly I can't provide the source code for the random number algorithm as it is proprietary, however I have attached the B4J test program which include randomNoGen.exe in the objects directory.
Hope you find this useful.
Best regards
Rob