plot with par(mfrow=c(3,2))

Question

I have data of 6 variables in excel (each variable in one column). But the length of each column is different. I want to plot this data by par(mfrow=c(3,2)) to divide the plot in 6 parts and plot all plots together ,but using the name of variable for each graph like the pic I attached. The variable names are: MKT SMB HML RMW CMA MOM And the plot code is

MEplot(S,main = "variable name",cex.main=2,bty="n",cex.axis = 1.3,xlab = "", ylab = "")

I can plot the same photo, but I have to repeat code 6 times because the column name is different, so each time I have to specified the range of column and repeat the code. Could you please help? How can I plot this graph with one code using for loop?

Are you actually using a function called `MEplot()`? We can’t tell because you didn’t include enough information for a [minimum reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example). As the link suggests, you should always provide any other `library()` that is part of your code. — John Polo, Dec 17 '22 at 11:07

Chris Ruehlemann · Answer 1 · 2022-12-17T14:35:06.663

Here's one approach using base R:

Set the plot parameters:
```
par(mfrow = c(2,3))
```

Perform for loop for each variable:

for(i in 1:6){
  plot(df[,i], main = names(df)[i], add = TRUE)
}

Refine plot parameters as needed inside the for loop. If you want to have same-scaled y-axes, use ylim = range(df).

Here's another approach using ggplot2 (has the advantage that the y-scales are automatically aligned)

library(ggplot2)
df %>%
  pivot_longer(everything()) %>%
  ggplot(aes(x = "", y = value)) + 
  geom_point() +
  facet_grid(~ name)

If you want your 2 by 3 layout use facet_wrap(~ name, nrow = 3)(thanks to @r2evans!)

Data:

df <- data.frame(
  MKT = rnorm(100),
  SMB = rnorm(100,2),
  HML = rnorm(100,5),
  RMW = rnorm(100,60),
  CMA = rnorm(100,33),
  MOM = rnorm(100,100)
)

Since you're introducing `ggplot2` here, a more representative example (mimicking the original `3x2` plot) could be `facet_wrap(~ name, nrow = 3)`. — r2evans, Dec 17 '22 at 12:09

jay.sf · Answer 2 · 2022-12-17T13:52:34.230

Using lapply.

par(mfrow=c(3, 2))
lapply(names(dat), \(x) plot(dat[[x]], main=x, xlab='', ylab='', ylim=range(dat)))

If you want to fill the layout column-wise, use mfcol= isntead of mfrow=. The ylim=range(dat) cares for the y axes to be in the same range.

Data:

set.seed(42)
dat <- setNames(as.data.frame(replicate(length(x), rnorm(100))), c("MKT", "SMB", "HML", "RMW", "CMA", "MOM"))
dat[] <- lapply(dat, sort)

Alternatively to par you can use a layout matrix.

layout(matrix(1:6, ncol=2, byrow=TRUE))
lapply(names(dat), \(x) plot(dat[[x]], main=x, xlab='', ylab='', ylim=range(dat)))

plot with par(mfrow=c(3,2))

2 Answers2