可以通过对 OPENCPU 的 API 调用直接运行 R 片段吗？

Question

I am trying to use opencpu server to run R code via API calls. For R Markdown it works fine (as explained in opencpu example markdown app). I want to directly pass R code in to opencpu and get output. Is that possible?

Some thing like http://localhost/ocpu/.../something {"text":"R code here"}

Answer 1

Yes, it is possible and well documented on the OpenCPU website, here

Argument formats for R function calls (HTTP POST only)

When calling a function, we need to pass arguments. OpenCPU accepts the following types of arguments: Argument:plain R code --- Example: paste("useR", "2013"), mtcars

Answer 2

I have found three ways to do this:

1. Post a .r file to OpenCPU at ../library/base/R/source
2. Use do.call
3. Create your own R package and save on the server

I'll demo the first two methods:

Method 1

Post a file with parameter name file to ../library/base/R/source

Here is an example, in the image below you can see I am making a POST request using localhost:8004/ocpu , where I have OpenCPU monitoring port 8004 in a Docker container.

Above, I inserted the file test.r. This is the contents of test.r:

g = "hello world"
x = "yoyoi"

So it is just a two line file. You can see OpenCPU returns the location parameter:
Location: http://localhost:8004/ocpu/tmp/x0ec7fa2f50f510/

If you go to that url, you will see the following:

console
files/DESCRIPTION
files/test.r
info
R/.val
R/source
source

R/.val is the output R object. So if you send get request to the link http://localhost:8004/ocpu/tmp/x0ec7fa2f50f510/R/.val , you will see the output of running file test.r :

$value
[1] "yoyoi"

$visible
[1] FALSE

So basically, you can send a get request to the location url returned from your post request. Then you will get the last line of test.r

Method 2

You can use do.call from package base to execute a function:

Above will return onetwo

More Detailed Example:
Here is a more detailed example. Since data persists on the OpenCPU server for a time, so you can use do.call in base R to execute a function on the output of utils::read.csv :

Step 1: read a file with read.csv, in this example I'm reading the cars dataset from here

Step 2: Apply do.call on the dataframe returned by read.csv :

This outputs the first 6 rows of the dataframe:

                  X  mpg cyl disp  hp drat    wt  qsec vs am gear carb
1         Mazda RX4 21.0   6  160 110 3.90 2.620 16.46  0  1    4    4
2     Mazda RX4 Wag 21.0   6  160 110 3.90 2.875 17.02  0  1    4    4
3        Datsun 710 22.8   4  108  93 3.85 2.320 18.61  1  1    4    1
4    Hornet 4 Drive 21.4   6  258 110 3.08 3.215 19.44  1  0    3    1
5 Hornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2
6           Valiant 18.1   6  225 105 2.76 3.460 20.22  1  0    3    1

You can also return as JSON:

http://localhost:8004/ocpu/tmp/x0ecb9593c335da/R/.val/json

I also just ran a test where I uploaded the cars data twice, which returned two different keys. I then used the key::.val notation with do.call to rbind the two dataframes together. So if you use that method you can even have an app where you upload 10 csv files and rbind them all together.