So when it comes to graphics file devices there are basically two broad categories of devices. One is called vector devices or vector formats and the other format is called bitmap. So vector formats are used, are most useful for things like line graphics. So, if you have a scatter plot and maybe you draw a regression line through it these are line graphics. They don't have, they're not like natural scenes or images of photographs or things like that. And so, these types of graphics are useful, are, are usefully kind of generated using vector formats because they resize well. So if you have to take the image and kind of make it bigger or smaller, that it won't get distorted and the, and the quality of the graphic won't be, won't suffer. For example the PDF format is very portable, you can use it on almost any platform that's out there right now. And so it is a good format, a general purpose vector format. Vector formats are typically not very good if you have a plot that has like, that has a lot, a large number of points because every single point has to be represented by a piece of information in the, in the, in the file format. And so, these files tend to be extremely large if you plot a lot of different points to a vector format file type. [INAUDIBLE] scalable vector graphics or SVGs, is a popular format for web based plots. It's useful because it supports things like animation and interactivity. In the plot and, and almost all web browsers recognize the SVG format now. Windows metafile is the, is, is the vector format that's available only on Windows. And you may see older files lying around that are in the post script format. This is kind of a predecessor to the PDF format, it's not used very often, but you may see it around. In the bitmap format file file formats there are few commonal ones. Probably the most common is PNG which stands for portable network graphics. The night, and so basically, bitmap formats are represent plots or images as a series of pixels, essentially. And so They are, they are very good for for [UNKNOWN] that had a lot of different points were, because they will be very efficiently represented in a bitmap format. the, a PNG is particularly good for things like line drawings or images with solid colors it uses a lossless compression algorithm. So that the file sizes tend to be small and, and pretty much every web browser can read PNGs now. So if you, so they're good for web, web based plots. JPEG you might be familiar with. If you have a digital camera or something like that. They're very good for things like natural scenes. Which don't, which don't have solid colors but have lot's of gradients of colors. It uses a lossy compression algorithm. So the file sizes are very efficient, very small and they're also good for just because it's a bitmap format, it's good for plotting, making plots with lots of points. Bitmap formats in general do not resize well, so when you create a plot and you try to make it bigger or smaller they can distort the quality of the image. And so generally speaking, you should not resize bitmap plots after they've been generated. JPEGs of course, can be read by any computer or in any web browser so they're very useful for that kind of, for web based plots. they, they're not very good for line drawings because if you create a line drawing in PD, in JPEG otherwise you'll see some aliasing in the lines. TIFF is another older format for for bitmap files. It supports lossless compression and it's and it's a very commonly used still. Bmp is a negative, a bitmap format for Windows. And is commonly used for things like icons. So it's possible to open multiple graphics devices. And so for example, for example, you might want to create three or four different plots to look at, at the same time, and the way to do that is to have say three or four different screen devices open. And so the way you can do this is, is to, is you can launch graphic devices explicitly you launch the screen device for example on Mac by calling the quarts function multiple times and then you'll open multiple graphic devices screen devices. And so but you can only plot to one device at a time. And so, the graphics device that you plot to is the active graphics device and you can figure out which graphics device is the cur, is the active one by calling the dev.cur() function or dev.cur stands for current. And that function will return an integer that will tells you which graphics device is currently active. So, every graphics device is as, is assigned an integer that's between two and and two and up basically. So there's no graphics device that has the number one. So, every graphics device gets an integer and you can change which graphics device is the active graphics device by calling the dev.set function. And dev.set takes an integer value which will switch you to the graphics device that is, that corresponds with the integer that you're given. So you can switch back and forth between the graphics devices using the dev.set function if you need to. Finally it's possible to copy plots from one device to another. And this is probably most commonly used when you've created a plot on the screen and you've decided that you really like it and then you want to, then you want to put it into a file. And so there's two ways that you can do this. One is that you can, after you, if you've saved the code. That can make the plot on the screen. You can just open up a file device and then cut and paste the code into R and then close the file device and then you got your plot in a file. But, a faster way and sometimes more convenient is to copy the plot from the screen device to the file device. And you can use this with the dev.copy function which just copies plots from one device from any device to any other device And then, if you specifically want a PDF file there is a function called dev.copy2pdf which will take the screen device and copy it directly to a PDF file. And it's important to note that copying the plot is not an exact operation so, the, the, the plot you end up with in the file, may not look exactly like the copy, the plot you saw on the screen. So it's a, there may be some adjustment you have to do. So here is some code that just creates a plot on the screen. This is the Old Faithful Geyser data that comes with R and then I want to copy it to a PNG file, so I want to create a PNG file on my computer that contains the plot that I'm looking at on my screen. So, I can call dev.copy, the first argument to dev.copy is the, is the function that opens the file device in this case, this is the PNG function. And then, I just give it a file name, which is needed to create the file. So, I just call it geiserplot.png. And then, once I've copied the plot, I still need to close the PNG device with dev.off. So, once I've call dev.off, there'll, there should be a file on my computer that, that's going to be in a PNG format. And I can send it to someone or include it in a presentation. So, that's a quick summary of the the graphics devices in R. The basic bottom lines are that every plot must be created on a graphics device. And usually this is the screen device, but and it's useful for exploratory analysis. But if you want to preserve a plot or send it to someone, you have to create a plot in a file device. And there a variety of different file formats that you can use. There are vector formats like PDF and SVG, and therebBitmap formats like PNG and JPEG. That you can chose from.
