Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: decay rate for univariate time series data


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: decay rate for univariate time series data
Date   Fri, 8 Jun 2012 19:44:00 +0100

"diff" == different, difficult, diffident?

Look like a Poisson regression to me. I don't see anything
intrinsically econometric about this either.

Nick

On Fri, Jun 8, 2012 at 7:30 PM, tashi lama <ltashi32@hotmail.com> wrote:

>   Of all the questions I have posted, this one is very diff since this is more of a econometrics problem than stata. I am hoping to get some thoughts and ideas. I have a univariate time series data where the variable is readership hits on a document over time. The problem is to identify the time till the document is still fresh. My goal is to find the "decay rate" as a metric to evaluate freshness of the document. Here is my dataset
>
>
>
>        day      hits
>     |------------------|
>  1. | 03jan2011    211 |
>  2. | 04jan2011     60 |
>  3. | 05jan2011     28 |
>  4. | 06jan2011     16 |
>  5. | 07jan2011     20 |
>     |------------------|
>  6. | 08jan2011      2 |
>  7. | 09jan2011      8 |
>  8. | 10jan2011     10 |
>  9. | 11jan2011      7 |
>  10. | 12jan2011     12 |
>     |------------------|
>  11. | 13jan2011      7 |
>  12. | 15jan2011      3 |
>  13. | 16jan2011      3 |
>  14. | 17jan2011      5 |
>  15. | 18jan2011      1 |
>     |------------------|
>  16. | 20jan2011      1 |
>  17. | 21jan2011      1 |
>  18. | 24jan2011      5 |
>  19. | 25jan2011      4 |
>  20. | 26jan2011      4 |
>     |------------------|
>  21. | 28jan2011      1 |
>  22. | 31jan2011      2 |
>  23. | 01feb2011      1 |
>  24. | 02feb2011      2 |
>  25. | 04feb2011      2 |
>     |------------------|
>  26. | 05feb2011      2 |
>  27. | 07feb2011      2 |
>  28. | 08feb2011      2 |
>  29. | 09feb2011      1 |
>  30. | 14feb2011      1 |
>     |------------------|
>  31. | 15feb2011      1 |
>  32. | 16feb2011      1 |
>  33. | 24feb2011      1 |
>  34. | 25feb2011      1 |
>  35. | 03mar2011      1 |
>     |------------------|
>  36. | 06mar2011      2 |
>  37. | 11mar2011      2 |
>  38. | 19mar2011      2 |
>  39. | 29mar2011      2 |
>  40. | 07apr2011      3 |
>     |------------------|
>  41. | 22apr2011      1 |
>  42. | 29apr2011      1 |
>  43. | 06may2011      2 |
>  44. | 07may2011      1 |
>  45. | 09may2011      2 |
>     |------------------|
>  46. | 16may2011      1 |
>  47. | 04jul2011      2 |
>  48. | 06jul2011      1 |
>  49. | 30jul2011      2 |
>  50. | 11aug2011      1 |
>     |------------------|
>  51. | 21sep2011      1 |
>  52. | 25oct2011      1 |
>  53. | 10nov2011      1 |
>     +------------------+

> My questions are:
>
>   1. Can I run a regression to get the dacay rate and if I can, what kind of regressions should I consider running? The scatter plot of readership hits looks like a negative exponential.
>
>   2. Please note my date has gaps. In other words, it is not continuous.
>
>   3. Any ideas to obtain a decay rate
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index