You may want to use -ipolate-.
Example:
clear
input Day Y_observed Y_Inferred
1 10 10
2 . 25
3 50 50
4 . 50
5 50 50
6 . 40
7 . 30
8 20 20
9 40 40
end
ipolate Y_observed Day , generate(Y_ipolate)
list
+--------------------------------------+
| Day Y_obse~d Y_Infe~d Y_ipol~e |
|--------------------------------------|
1. | 1 10 10 10 |
2. | 2 . 25 30 |
3. | 3 50 50 50 |
4. | 4 . 50 50 |
5. | 5 50 50 50 |
|--------------------------------------|
6. | 6 . 40 40 |
7. | 7 . 30 30 |
8. | 8 20 20 20 |
9. | 9 40 40 40 |
+--------------------------------------+
(Note that your value for Y_Inferred for Day 2 is probably wrong).
HTH,
Philipp
Inamo, Jocelyn F. M.D., Ph.D. wrote:
Dear STATA Users
I would like to fill the missing values in my time series data shown in
the first table, assuming that there is a linear progression between the
observed proxy values. Accordingly, I would like to obtain inferred
values of Y as shown in the second table.
X Y
1 10
2
3 50
4
5 50
6
7
8 20
9 40
Day Y observed Y Inferred
1 10 10
2 25
3 50 50
4 50
5 50 50
6 40
7 30
8 20 20
9 40 40
The most accurate way to do this would be to go to each missing value
and to compute a linear equation within the proxy values of Y. I
realized that it was more easy to say than to do (for me at least).
The second way would be to use the smoothing commands of stata. I
realized that ksm-lowess work pretty well when set with the smallest
bandwidth possible, but I have been unable to recover the inferred
values of Y.
Any suggestion?
Thanks
Jocelyn
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
