CODEWITHSUNDEEP
r
math
math

Reputation: 2022

Some basic question in R writing a R-script for a regression example

First of all I want to point out that I'm not very familiar with R, so sorry if one of the following questions is clear.

My motivation is to write a simple R-script, which should contain:

  1. import data
  2. do regression of form $ Y=aX+bZ+intercept$
  3. some calculations
  4. ouput

now here are my questions:

  1. This is a very general question: If I wrote the R script, then I have to load it with source(name.R), right? Must be there an additional command to execute the script?
  2. Suppose I did my regression with lm, like fit<-lm(Y~X+Z,data=database) this gives a nice ouput. What I really want is to save the coefficients of the model in a vector. How can I do this? Here would it be a 3-dimensional vector (intercept, a, b). EDIT I've tried coefficient<-coefficient(fit). This does not work! coefficient is not a numerical vector. There are also the name, i.e. intercept and the value below for the first element of it.
  3. If I want to print out the coefficients and some calculations at the very end of the script, how do I do this? Just write print(....)?

I'm very thankful for your help and Hopefully I considered all rules and conventions in this forum, since this is my first question. If not, I'm very sorry.

Upvotes: 0

Views: 528

Answers (1)

nico
nico

Reputation: 51670

If I wrote the R script, then I have to load it with source(name.R), right? Must be there an additional command to execute the script?

Not if your script directly invokes the commands

For instance if name.R contains

a <- 1:10
plot(a, a^2, t="l")

Then source("name.R") will directly generate a plot

However, if name.R contains

myfunction <- function()
   {
   a <- 1:10
   plot(a, a^2, t="l")
   }

Then sourcing it will only load the function. You will then have to invoke myfunction() to get the plot.

Suppose I did my regression with lm, like fit<-lm(Y~X+Z,data=database) this gives a nice ouput. What I really want is to save the coefficients of the model in a vector. How can I do this? Here would it be a 3-dimensional vector (intercept, a, b)
If I want to print out the coefficients and some calculations at the very end of the script, how do I do this? Just write print(....)?

 print(coef(fit))

will give you what you need (you can store them in an array with model.coef <- coef(fit))

Also, it can be interesting to run

 summary(fit)

See ?coef and ?summary for more info

Upvotes: 2

Related Questions