How to call the par(), settings embed an object in R? -

before considering answer, correct question: , regarding question: “ i'm curious why, though, setting parameters after plotting vice before? “ actually, forgot mentioning something. correct order this:

using data again:

    dose <- c(20, 30, 40, 45, 60)      druga <- c(16, 20, 27, 40, 60)      drugb <- c(15, 18, 25, 31, 40) 

1st:run: plot(dose, druga, type="b", pch=19, lty=2, col="red" (it self checking mechanism. mean proving myself first, generating plot default r settings)(plot_1)

2nd_run: opar.1<-par(pin=c(5, 3), lwd=2, cex=1.5)

3rd_run: plot(dose, druga, type="b", pch=19, lty=2, col="red") (same plot opar.1 setting, compare first plot , second plot)(plot_2)

4th_run: par(opar.1) (with intention of resetting opar.1 settings.)

5th_run: plot(dose, druga, type="b", pch=19, lty=2, col="red") (and seeing third plot same first plot. hence, proves par(opar.1) command reset opar.1 settings , turn default settings.)(plot_3 same plot_1)

but, 5th_ when run opar.1 , 6th_run plot(dose, druga, type="b", pch=19, lty=2, col="red", not getting plot same plot_2. in conclusion,i wondering if getting wrong thinking “opar.1<-par(pin=c(5, 3), lwd=2, cex=1.5)” command store “par(pin=c(5, 3), lwd=2, cex=1.5)” settings , if use “par(opar.1)” command after “opar.1<-par(pin=c(5, 3), lwd=2, cex=1.5)” command , reset opar.1 settings, running “opar.1” can load “par(pin=c(5, 3), lwd=2, cex=1.5)” settings?

there 2 mechanisms come mind, , though starting in right direction, you're missing final point(s). i'm curious why, though, setting parameters after plotting vice before? useful, typically need set them before calling plot.

using data:

dose <- c(20, 30, 40, 45, 60) druga <- c(16, 20, 27, 40, 60) drugb <- c(15, 18, 25, 31, 40) 

1. the first mechanism not require additional packages:

   opar.1 <- par(pin = c(5, 3), lwd = 2, cex = 1.5) plot(dose, druga, type = "b", pch = 19, lty = 2, col = "red")  par(opar.1) 

   one risk of doing way that, before restorative call par, if code fails (causing stop or other early-exit function), second par command may not called. in function (not command-line), problem can avoided following, provided (as @rawr mentioned) in ?par:

   opar.1 <- par(pin = c(5, 3), lwd = 2, cex = 1.5) on.exit(par(opar.1)) plot(dose, druga, type = "b", pch = 19, lty = 2, col = "red")  

   (this want when executed in function ... on command-line, you'll need first instance of calling par directly.)

2. using withr package:

   library(withr) message('## before with_par') # ## before with_par str(par(c('pin', 'lwd', 'cex'))) # list of 3 #  $ pin: num [1:2] 5.76 5.16 #  $ lwd: num 1 #  $ cex: num 1 with_par(list(pin = c(5, 3), lwd = 2, cex = 1.5), {   plot(dose, druga, type="b", pch=19, lty=2, col="red")   message('## inside with_par')   str(par(c('pin', 'lwd', 'cex'))) }) # ## inside with_par # list of 3 #  $ pin: num [1:2] 5 3 #  $ lwd: num 2 #  $ cex: num 1.5 message('## after with_par') # ## after with_par str(par(c('pin', 'lwd', 'cex'))) # list of 3 #  $ pin: num [1:2] 5.76 5.16 #  $ lwd: num 1 #  $ cex: num 1 

   the message , str calls before, inside, , after with_par demonstrate parameters changed inside code block, relatively safe both in functions , on command-line.

both methods provide identical par-adjusted plots.