1. ## Random Numbers (na)

This isn't strictly an excel question - it is more a random number question but I wasn't sure where to put it.

I have 5 empirical distributions from which I want to randomly select values. I have generated a long long series of random numbers from a suitable random number engine. I then split this series into 5 sub-series in order, thus

sub-series 1 takes observations number (from long series):- 1, 6, 11, 16, 21, 26, etc
sub-series 2 takes observations number (from long series):- 2, 7, 12, 17, 22, 27, etc
sub-series 3 takes observations number (from long series):- 3, 8, 13, 18, 23, 28, etc
sub-series 4 takes observations number (from long series):- 4, 9, 14, 19, 24, 29, etc
sub-series 5 takes observations number (from long series):- 5, 10, 15, 20, 25, 30, etc

The question is: are the sub-series random? Or should I really be generating 5 separate series from the engine?

2. ## Re: Random Numbers (na)

Tim,

I think that any subseries of a truely random number series should also be random. Before the days of computers, people would extract blocks from random number tables in this manner. Maybe the online Randomizer would help?

Alan

3. ## Re: Random Numbers (na)

So long as your method of sub-set selection is not related to the numbers in any way, a sub-set of random numbers is also random.

The issues are more likely to relate to sample-size, than randomness. There are also tests for randomness that could be applied to the subsets - these would be likely to flag that your sub-sets were too small (if they were) - but of course, that may be the purpose of your sub-setting <img src=/S/grin.gif border=0 alt=grin width=15 height=15>. e.g. tonight's lotto numbers may not pass a randomness test on their own, but they are random numbers.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•