如何在Python接口中运行R代码？

Question

我试图运行下面的代码，我想读取csv文件，然后编写"sas7bdat“。我试过了下面的代码。

我们已经在系统上安装了R.

from rpy2 import robjects    

robjects.r('''
        library(haven)
        data <- read_csv("filename.csv")
        write_sas(data, "filename.sas7bdat")
        ''')

在运行上面的代码之后，就不会有由这些代码生成的输出，甚至我也不会得到任何错误。

期望输出：试图读取.csv文件，然后以.sas7bdat格式导出数据。(在标准python 3.9.2编辑器中)

python没有这样的功能/库，因此我正在尝试以.sas7bdat格式导出数据。

请建议对上述代码或python中的任何其他方式进行一些更改，通过这些方式，我可以在python中创建/导出.sas7bdat格式。

谢谢。

Answer 1

我在Python木星笔记本中有使用R的经验，刚开始的时候有点复杂，但确实起作用了。在这里，我刚刚贴上了我的个人笔记，希望能帮上忙：

# Major steps in installing "rpy2":
# Step 1: install R on Jupyter Notebook: conda install -c r r-essentials
# Step 2: install the "rpy2" Python package: pip install rpy2 (you may have to check the version)
# Step 3: create the environment variables: R_HOME, R_USER and R_LIBS_USER 
# you can modify these environment variables in the system settings on your windows PC or use codes to set them every time)

# load the rpy2 module after installation
# Then you will be able to enable R cells within the Python Jupyter Notebook
# run this line in your Jupyter Notebook
%load_ext rpy2.ipython

我的工作是用Python做ggplot2，所以我做了：

# now use R to access this dataframe and plot it using ggplot2
# tell Jupyter Notebook that you are going to use R in this cell, and for the "test_data" generated using the Python
%%R -i test_data 
library(ggplot2)

plot <- ggplot(test_data) + 
        geom_point(aes(x,y),size = 20)
plot
ggsave('test.png')

Answer 2

在运行代码之前，请确保haven和reader安装在R kernel中。

from rpy2.robjects.packages import SignatureTranslatedAnonymousPackage

string = """
         write_sas <- function(file, col_names = TRUE, write_to){
             
             data <- readr::read_csv(file, col_names = col_names)
             haven::write_sas(data, path = write_to)
 print(paste("Data is written to ", write_to))

}
""" 

rwrap = SignatureTranslatedAnonymousPackage(string, "rwrap")

rwrap.write_sas( file = "https://robjhyndman.com/data/ausretail.csv", 
                col_names = False,
                write_to = "~/Downloads/filename.sas7bdat")

您可以使用任何R函数参数。和我用col_names一样