Preface

To be clear, I have done something very useless. (I will put it for saving the blunt code) I wanted to make it more convenient to use python in Rstudio, so I made a function, but in the end, the convenience has hardly improved. ..

Click here for how to use python in Rstudio.

Python dataframe

Thanks. I love reticulate.

It is common to use pandas for python data frames, but there is a problem that it is difficult to see the data frames when using python with Rstudio.

For example,

#"Data Scientist Association Skill Definition Committee"
#Data of "Data Science 100 Knock (Structured Data Processing)"
#I will use it.

>>> import pandas as pd
>>> df = pd.read_csv("customer.csv")
>>> print(df)

名称未設定.png

Well, it's not bad. .. It's been omitted on the way, so it's a little scary!

Rstudio View () function

The solution is simple.

・ Use google colab obediently (The dataframe looks relatively nice)

-** Use Rstudio-specific View () function **

The View () function is the quickest.

>>> quit #Exit python and return to R

> # library(reticulate)
> View(py$df)

This will bring up a super easy-to-read dataframe that can be moved as usual. 名称未設定.png

Functionalization

Then this

"Loading the reticulate package (py $ cannot be used unless loaded)-> View in Viewer-> Return to python"

Let's make a series of actions like a function.

py_view = function(py_df_name, i_overwrite = FALSE, repl = TRUE){
  library(reticulate)
  #
  if(class(py_df_name) == "character"){
    if((class(try(py$i, T)) == "try-error") |
       (class(try(py$i, T)) != "try-error" & i_overwrite == TRUE)){
      try(py_run_string(paste("i = globals()['", py_df_name, "']",
                              sep = "")), T)
      if(class(try(class(py$i), T)) == "data.frame"){
        View(py$i, py_df_name)
        py_run_string("del i")
        #repl
        if(repl == T){repl_python()}
      }else{
        message(paste("Data frame'", py_df_name, "'Was not found.",
                      sep = ""))
        try(py_run_string("del i"), T)
      }
    }else{
      message("   'i'Exists. Delete or argument i_overwrite =Please set TRUE.")
    }
  }else{
    message("Argument py_df_Please specify the name as a character string.")
  }
}

I had a hard time. First of all, if you set the argument (dataframe name) to "py $ ~", it will end in an instant, but I made it a form to bring in a dataframe with a name that matches in the character string.

The behavior is that if the argument py_df_name is a character (string), move the dataframe to a variable called i in python and look at it with the View () function.

You've created a number of exception handling by try. Commentary ...

Try using

Easy to use.

> py_view("df")
>>>

With this alone, the previous Viewer is displayed and you can return to python.

Summary

that's all! !! !! It was more wasteful than I expected.

I thought it would be annoying to have to add "py" every time in View () and to stumble with "py $ ~" if I didn't do "library (reticulate)", but in the end I made this I think it's faster to work.

For the time being, it was good practice for exception handling in R.