filling missing data: polynomial spline

[Des]

Hello all,

I am new to this forum, and hope someone will step up helping out on the following.

I have a longitudinal data, but with some missing values of individuals' weight.   If  at least 50% of the data points are not missing, I want to fill missing values based on the overall pattern of observed values (polynomial spline).

data have;

infile cards;

input ID time wt;;

cards;

1      1      54

1      2      66

1     3       .

1     4       81

1     5      82

1     6      .

1     7      .

1     8      .

1     9     67   

1     10    64

2    1     75

2     2     72

2    3      .

2    4      .

2    5     .

2    6     .

2    7      80

2    8      79

2    9      76

2    10      84

;

Can any help me in coding to fill these missing values with polynomial spline instead of nearest values?

[PGStats]

Do you have the license to SAS/ETS? PROC EXPAND does exactly what you want.

PG