> mdata <-read.table("seisan.csv", header=T, sep=",")
> ly <-log(mdata[,1])
> lk <-log(mdata[,2])
> ll <-log(mdata[,3])
> seisan_result <-lm(ly~lk+ll)
> summary(seisan_result)

Call:
lm(formula = ly ~ lk + ll)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.42905 -0.17434 -0.06668  0.16183  0.65620 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  0.11545    0.71209   0.162    0.873    
lk           0.56779    0.07812   7.268 3.70e-07 ***
ll           0.51940    0.08534   6.086 4.86e-06 ***
---
Signif. codes:  0 �e***�f 0.001 �e**�f 0.01 �e*�f 0.05 �e.�f 0.1 �e �f 1

Residual standard error: 0.2777 on 21 degrees of freedom
Multiple R-squared:  0.9518,    Adjusted R-squared:  0.9473 
F-statistic: 207.5 on 2 and 21 DF,  p-value: 1.472e-14

> library(car)
> linearHypothesis(seisan_result,"lk+ll=1",test="F")
Linear hypothesis test

Hypothesis:
lk  + ll = 1

Model 1: restricted model
Model 2: ly ~ lk + ll

  Res.Df    RSS Df Sum of Sq      F Pr(>F)
1     22 1.8200                           
2     21 1.6192  1   0.20081 2.6043 0.1215