Groups | Search | Server Info | Login | Register

Groups > comp.soft-sys.sas > #969

Re: interpolating missing values

Show all headers | View raw

Sorry about the multiple posts, had some trouble submitting.
I know that this can be done in SAS, my question is how?
The most difficult part for me is how to reference the first
observations before and after a set of blank values to use in the
interpolation.

Bryan




On Jun 20, 6:22 pm, "Kenneth M. Lin" <kenneth_m_...@sbcglobal.net>
wrote:
> You don't have to post the same question four times.  Yes, you can indeed
> batch interpolates in SAS but you will have to write your own code.  For
> starter, for each observation you could obtain the relevant information you
> would need to perform the interpolation.  And when the observation is
> missing, you replace it with the calculated value.
>
> "wyldsoul"  wrote in message
>
> news:4be0f2da-c05f-40ba-95a7-cda03c9e3854@s16g2000prf.googlegroups.com...
>
> Hi all,
> I want to interpolate missing values from gage data.  This is easy to
> do manually in excel, however, I have over 60 years of water level
> data from dozens of stage gages, and would like to find a way to batch
> interpolate in SAS.   In order to interpolate, for each set of missing
> values I need to  calculate the difference between the last
> observation before the first blank value and the first observation
> after the last blank value.  Then divide the result of this by the
> total number of missing values+1.  Simply put:
> (first obs± last obs)/(# missing obs+1)
> or using the sample data from below: (6.05-6.01)/4=0.01
>
> Then I need to populate each missing observation by cumulatively
> adding or subtracting the result of the previous calculation to each
> successive blank until you reach the next non-blank observation.
> The sample dataset is below.  Stage is the original value, int_factor
> is the result of the first calculation, and interp_stage is the
> interpolated stage based on cumulatively adding the int_factor to each
> observation .
>
> Station date stage int_factor interp_stage
> NP-P33 1-Jan-53 6.04 6.04
> NP-P33 2-Jan-53 6.03 6.03
> NP-P33 3-Jan-53 6.02 6.02
> NP-P33 4-Jan-53 6.01 6.01
> NP-P33 5-Jan-53 0.0100 6.02
> NP-P33 6-Jan-53 6.03
> NP-P33 7-Jan-53 6.04
> NP-P33 8-Jan-53 6.05 6.05
> NP-P33 9-Jan-53 6.15 6.15
> NP-P33 10-Jan-53 6.20 6.2
> NP-P33 11-Jan-53 6.20 6.2
> NP-P33 12-Jan-53 6.19 6.19
> NP-P33 13-Jan-53 6.18 6.18
> NP-P33 14-Jan-53 6.17 6.17
> NP-P33 15-Jan-53 6.16 6.16
> NP-P33 16-Jan-53 6.15 6.15
> NP-P33 17-Jan-53 6.14 6.14
> NP-P33 18-Jan-53 6.13 6.13
> NP-P33 19-Jan-53 6.12 6.12
> NP-P33 20-Jan-53 0.0091 6.11
> NP-P33 21-Jan-53 6.10
> NP-P33 22-Jan-53 6.09
> NP-P33 23-Jan-53 6.08
> NP-P33 24-Jan-53 6.07
> NP-P33 25-Jan-53 6.07
> NP-P33 26-Jan-53 6.06
> NP-P33 27-Jan-53 6.05
> NP-P33 28-Jan-53 6.04
> NP-P33 29-Jan-53 6.03
> NP-P33 30-Jan-53 6.02 6.02
> NP-P33 31-Jan-53 6.02 6.02

Back to comp.soft-sys.sas | Previous | Next — Previous in thread | Find similar

Thread

interpolating missing values wyldsoul <wyldsoul@gmail.com> - 2011-06-20 13:25 -0700
  Re: interpolating missing values "Kenneth M. Lin" <kenneth_m_lin@sbcglobal.net> - 2011-06-20 15:22 -0700
    Re: interpolating missing values wyldsoul <wyldsoul@gmail.com> - 2011-06-21 07:03 -0700

csiph-web