![graphpad prism 6 add data labels graphpad prism 6 add data labels](https://windows-cdn.softpedia.com/screenshots/GraphPad-Prism_1.png)
annotate() is quite limited compared to what I mean. Iirc, it interacted strangely with previously defined aesthetics. I've also just had it flat-out fail for reasons the devs couldn't explain, and didn't seem to care about. Stat_function() can't be faceted (base R has no problem with this, since you do faceting yourself).
![graphpad prism 6 add data labels graphpad prism 6 add data labels](https://cdn.graphpad.com/assets/0.45.1/images/srcset/prism-organize-your-data-w1920-800.png)
I said that I don't have a link for complex plots, but I have created many in the past that were not feasible in ggplot (or, at least, if they were, they would require many, many more lines of code). The point of the article was moreso to show that 'what ggplot can do, base can also do'. But ggplot doesn't produce *everything* you could make by hand, so base R can do more. Since ggplot always produces something you *could* make by hand, base R graphics can do anything ggplot does. ggplot is *not great* for times when you just want to build a plot as you would by hand (like plotting arbitrary functions, adding points manually, and annotating). Basically, ggplot is *great* for taking data in a data frame, and mapping it to graphical things (literally, what ggplot is built on). I find custom plots obnoxious in ggplot for this reason - Sometimes I just want to feed it coordinates or a function, but that usecase is an afterthought at best.
![graphpad prism 6 add data labels graphpad prism 6 add data labels](https://www.graphpad.com/guides/prism/8/user-guide/images/frame18_zoom65.png)
In general, if you can draw it on paper, you can make it in base R that's not fully true in ggplot, because ggplot is fairly opinionated on what things can be mapped. ggplot, on the other hand, is a framework for expressing 'what' things are mapped to, and you just hope ggplot can resolve it into what you would draw. base R is really manual-plotting: "how I would draw it on paper, in order" it doesn't need a data frame, it doesn't need expressions, it only needs instructions and coordinates. What immediately comes to mind is plotting functions and adding custom annotations. I do know in my own years of using R, I've hit several problems that were far and away easier in base R than ggplot2 for creating a somewhat 'complex' plot. Are those easily created using R?īut I don't have a link about complex base plots. Two more examples from figures I created using Prism is this figure which shows a correlation, but shows information about the different groups of the participants as white and black circles or this figure that shows the mean and standard deviation but also the individual responses in between. Do you think it's easy to put the figures side by side, put information such as the '14 days' on the X axis, the group interaction also including the line pattern (HI vs LI), and the statistics on top of the figures, using code? (those figures were created using Prism. Please don't misunderstand, I love R and I think it's amazing, I just want to make sure that what I need to do for my job is possible with another tool, so as to not invest a considerable amount of time but to keep having to use the old tools.įor example, this is an article I recently published. However, I'm not confident that all the fine-grained changes in a figure are possible with writing code. I would like to invest some time to start creating my plots in R.