多重绘图功能R

Question

I am new to R and have been trying to find a solution to this for the past week via google and forums. My problem: I have a data set which I need to plot against age. There are over a 1000 variables with different measurement during in 40 different conditions. Looks like this:

Age   Variables1  Variable2 (....) Variable1000 > 
 |        |
 |        |
 v        v

What I need to do is plot the condition(age) against each of the columns of variables and output as different plots (all of this is just scatterplots). What is more, I want the output to be limited to only those variables that have a positive trend line coefficient.

So currently I have this very ugly code that is essentially a rough draft of what I really need.

plotest <- function(lung){
  # need to add the condition of abline function coefficient > 0 before plotting    
  plot(lung$Age, lung$hsa.let.7a.1, xlab = "Age", ylab = "miRNA")
  abline(lm(lung$hsa.let.7a.1 ~ lung$Age), col= "red")
  return(plot)
}
par(mfrow=c(2,2))
for (i in lung{plotest(i)})

I know this is mostly wrong. So sorry for the horrendous everything about it.

Could anyone direct me to any sources, which I might have overlooked in how to specify ranges in such large datasets? And function grammar? I have done some Python but found R to be much more confusing in this regard...

Thanks all, Paul

Answer 1

This should come pretty close to what you're asking for, although what you're going to do with 1000 graphs is beyond me.

# make up some data
x <- seq(1,10,len=100)
set.seed(1)    # for reproducible example
df <- data.frame(x,y1=1+2*x+rnorm(100), 
                   y2=3-4*x+rnorm(100),
                   y3=2+0.001*x+rnorm(100))

# this does the work...
lapply(colnames(df)[-1],function(col){
  form <- formula(paste(col,"x",sep="~"))
  fit  <- lm(form,df)
  if (coef(fit)[2] >0) {
    plot(form,df)
    abline(fit)
  }
})

Your code was not that far off. This example takes all the column names except the first one ( colnames(df)[1] ) and passes them one at a time to the function. The function creates a formula variable using the column name and the name of the first column, calls lm(...) , checks that the coefficient of x is > 0, and if so plots the data and the best fit line.

Look up the documentation on formula(...) , lm(...) , and coef(...) . Note that this example has a variable, y3 with a slope that is positive, but not significantly different from 0. You should think about how you want to deal with that situation.