Comments on Simulating the Birthday Problem with data derived probabilitiesTypePad2012-06-04T05:08:12ZBlog Administratorhttp://blog.revolutionanalytics.com/tag:typepad.com,2003:http://blog.revolutionanalytics.com/2012/06/simulating-the-birthday-problem-with-data-derived-probabilities/comments/atom.xml/Derek O'Connor commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b017616aa70e6970c2012-07-24T09:44:42Z2012-07-24T21:23:24ZDerek O'Connorhttp://www.derekroconnor.netDoes Revolution R have a Birthday Problem? Here is the Matlab answer: http://www.mathworks.co.uk/matlabcentral/answers/879-does-matlab-have-a-birthday-problem http://www.scribd.com/doc/48243415/O-Connor-The-Birthday-Paradox-and-Random-Number-Generation<p>Does Revolution R have a Birthday Problem?</p>
<p>Here is the Matlab answer:</p>
<p><br />
http://www.mathworks.co.uk/matlabcentral/answers/879-does-matlab-have-a-birthday-problem</p>
<p>http://www.scribd.com/doc/48243415/O-Connor-The-Birthday-Paradox-and-Random-Number-Generation</p>David Smith commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b0167678d37e2970b2012-06-15T13:07:25Z2012-06-15T13:07:25ZDavid Smithhttp://www.revolutionanalytics.com@Chris -- thanks! I really liked your time series chart, which I reproduced in this post.<p>@Chris -- thanks! I really liked your time series chart, which I reproduced in <a href="http://blog.revolutionanalytics.com/2012/06/birthday-problem-ctd.html" rel="nofollow">this post</a>.</p>Chris Mulligan commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b01676743358b970b2012-06-09T20:26:04Z2012-06-10T17:53:24ZChris Mulliganhttp://chmullig.com/2012/06/births-by-day-of-year/The CDC data includes birth day for 1969 to 1988. I actually did a very similar simulation a few days...<p>The CDC data includes birth day for 1969 to 1988.</p>
<p>I actually did a very similar simulation a few days ago using that full data and found that they were nearly identical. Slightly more likely in reality than the simulation, but only 0.14% more likely at n=23 (and n=23 is still the minimum group size necessary for >= 50%. </p>
<p>Post: <a href="http://chmullig.com/2012/06/births-by-day-of-year/" rel="nofollow">http://chmullig.com/2012/06/births-by-day-of-year/</a></p>Rick commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b01761530b847970c2012-06-09T03:35:08Z2012-06-09T14:36:13ZRickWhere is the data? The CDC files linked only include Birth Month and Year, and not the day.<p>Where is the data? The CDC files linked only include Birth Month and Year, and not the day. <br />
</p>Eric commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b01630636fd6a970d2012-06-07T15:34:42Z2012-06-07T15:34:42ZEricLink to punkrockor blog post is broken.<p>Link to punkrockor blog post is broken.</p>Carl Witthoft commented on 'Simulating the Birthday Problem with data derived probabilities'tag:typepad.com,2003:6a010534b1db25970b0176151ebe91970c2012-06-07T12:20:46Z2012-06-07T12:20:46ZCarl WitthoftIs the Feb 29 valley adjusted by 4x to account for the rate of occurrence of Feb29s? Meanwhile, I do...<p>Is the Feb 29 valley adjusted by 4x to account for the rate of occurrence of Feb29s?<br />
Meanwhile, I do find it amusing that the birthrate is slightly lower (heatmap) on the 4th of July and Christmas and New Year's day. I bet that first dip applies only in the USA :-) .<br />
And, finally, a pedantic note: given that the calendar year is cyclic and the end of the year is selected arbitrarily, it's a bad idea to attempt a linear fit, especially after you've overlaid several years' data. What does the fit look like if you plot "true time," i.e. from day 1 of first year to day 365 of the last year?</p>