CODEWITHSUNDEEP
rcbind
Rik
Rik

Reputation: 1987

Column-by-Column cbind of lm output

I'd like to do a column-by-column comparison of two lm outputs. Something like this except I'd like to see the estimates next to the estimates nest to the estimates, the Std Error next to the Std Error, etc. 

    x=runif(6);y=runif(6);z=runif(6)
    x2=runif(6);y2=runif(6);z2=runif(6)
    lm1=lm(x~z+y)
    lm2=lm(x2~z2+y2)
    s=summary(lm1)[["coefficients"]]
    s2=summary(lm2)[["coefficients"]]
    > cbind(s,s2)
               Estimate Std. Error    t value  Pr(>|t|)   Estimate Std. Error   t value   Pr(>|t|)
(Intercept)  0.67359508  0.8137390  0.8277778 0.4684978  0.6944551  0.2398032  2.895938 0.06270911
z           -0.19911923  1.0491558 -0.1897899 0.8615892  1.2509814  0.5678805  2.202896 0.11486078
y           -0.07097291  0.4569228 -0.1553280 0.8864254 -0.8926959  0.4984554 -1.790924 0.17123009

Upvotes: 3

Views: 202

Answers (1)

zx8754
zx8754

Reputation: 56219

Try this:

# column bind
res <- cbind(s,s2)

# make unique column names
colnames(res) <- make.unique(colnames(res))

# sort column names
ix <- order(colnames(res))

# result
res[,ix]

# output

#               Estimate Estimate.1  Pr(>|t|) Pr(>|t|).1 Std. Error Std. Error.1    t value t value.1
# (Intercept)  1.7701957 0.09783934 0.5505218  0.7042956  2.6399847    0.2342630  0.6705325 0.4176475
# z            0.2393442 0.45691026 0.8111899  0.6096530  0.9180588    0.8041508  0.2607068 0.5681898
# y           -1.7020430 0.12906057 0.6185858  0.8798162  3.0758497    0.7846930 -0.5533570 0.1644727

Upvotes: 4

Related Questions