Econ 7010 - Assignment 1 Hints

---

This tutorial will go through all the code you need for assignment 1.

Simple arithmetic

You can perform arithmetic to saved objects:

a <- 2
b <- 3
c <- 4
a + b * c

## [1] 14

Load data

For an example dataset, I will load the Cobb-Douglas data directly from the internet:

mydat <- read.csv("http://home.cc.umanitoba.ca/~godwinrt/7010/cobbdouglas.csv")

Note that you can load any .csv dataset this way, by typing the location of the file in the quotations above.

Calculate a sample mean

The sample mean for my dependent variable ,log(output), is:

mean(log(mydat$output))

## [1] 9.770502

Note that the mydat$ part is telling R where to find the variable output.

Estimate and save a model

I will estimate the Cobb-Douglas production function, and save it as an “object”. I can choose any name for the object (I choose “mymod”):

mymod <- lm(log(output) ~ log(labour) + log(capital), data = mydat)

Use the estimated model

To see the results of the estimation you can use summary(mymod) to see lots of information, but if you just want to see the estimated \(\boldsymbol{\beta}\), run mymod:

mymod

## 
## Call:
## lm(formula = log(output) ~ log(labour) + log(capital), data = mydat)
## 
## Coefficients:
##  (Intercept)   log(labour)  log(capital)  
##       0.1768        0.2652        0.9121

To get, and save, the LS residuals from the estimated model, use:

myresids <- residuals(mymod)

To get the estimated coefficients from the model use:

coefficients(mymod)

##  (Intercept)  log(labour) log(capital) 
##    0.1768209    0.2651853    0.9121483

and to get only the estimate for labour (for example), extract the 2nd element:

coefficients(mymod)[2]

## log(labour) 
##   0.2651853

Estimation without an intercept

Usually, it’s a good idea to include an intercept in the model, but sometimes we don’t want it. R includes an intercept by default. To get rid of the intercept, put -1 at the end of the equation:

lm(output ~ labour + capital -1, data = mydat)

## 
## Call:
## lm(formula = output ~ labour + capital - 1, data = mydat)
## 
## Coefficients:
##  labour  capital  
##  42.914    1.621

(I omitted the log for simplicity).