[R] inter-timeseries correlation or corrections

[Tim Michelsen]

Dear R-Users,
I put a small sample data set and script.

Aim: combine two partly overlapping series to one by prediction.

Problem: only overlapping data points are predicted.

Question: How do I predict data for rows 1-9 and 14-16?

Thanks in advance for your advince,
Tim

### CODE ###

x <- read.csv('Buch1.csv')
x
    year    A    B
1  1960    2   NA
2  1961    3   NA
3  1962    4   NA
4  1963   54   NA
5  1964    3   NA
6  1965  324   NA
7  1966  546   NA
8  1967  455   NA
9  1968  456   NA
10 1969  456   67
11 1970 8987   34
12 1971   78   34
13 1972   65    3
14 1973   NA  123
15 1974   NA 7688
16 1975   NA  324

predict(lm(a~b))
       10       11       12       13
2527.198 2394.489 2394.489 2269.824

### DATA ###
year,A,B
1960,2,
1961,3,
1962,4,
1963,54,
1964,3,
1965,324,
1966,546,
1967,455,
1968,456,
1969,456,67
1970,8987,34
1971,78,34
1972,65,3
1973,,123
1974,,7688
1975,,324