1

I'm trying to write Python code in an Rmarkdown file in RStudio. Generally speaking it works fine. I can write code chunks and knit them perfectly. I can also import and use packages: anything using, say, numpy or math runs and knits well.

Despite being able to run and knit code chunks that use other packages, I can't seem to knit the file when it contains chunks that use pandas. I can run these chunks and obtain the expected output. A minimal example would be simply:

import pandas as pd

which returns the (expected) output:

Python 3.9.4 (/usr/local/bin/python3)
Reticulate 1.20.9000 REPL -- A Python interpreter in R.

(I can also run code chunks using pandas functions.)

The problem is that, when I knit, I get the error:

Error in py_call_impl(callable, dots$args, dots$keywords) : ModuleNotFoundError: No module named 'pandas' Detailed traceback: File "<string>", line 1, in <module> File "/Library/Frameworks/R.framework/Versions/4.0/Resources/library/reticulate/python/rpytools/loader.py", line 44, in _import_hook level=level Calls: <Anonymous> ... py_capture_output -> force -> <Anonymous> -> py_call_impl Execution halted

I don't understand why I am able to run these code chunks, but not knit them.

Thank you for your help!

Gil Henriques
  • 53
  • 4

1 Answers1

2

Does your Rmarkdown file look somewhat similar to this for a simple example?

---
title: "Untitled"
author: "User"
date: "5/6/2021"
output: html_document
---

    
```{r setup, include=FALSE}
library(reticulate)
```
    
```{python}
import pandas as pd
iris = pd.read_csv('https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv')
iris
```

And you still get error? sometimes you may need to reinstall pip packages, usually pip3 install pandas works, you can also add py_install("pandas") inside your r chunk under library(reticulate) to have it install pandas to help it render properly. Sometimes making simple examples help break the problem down into smaller chunks

Daniel_j_iii
  • 3,041
  • 2
  • 11
  • 27
  • Thank you so much Daniel! Running `py_install("pandas")` solved the issue. I had already installed the package using pip3, so it is not quite clear to me why this was necessary... (Also, it doesn't seem to be necessary for me to load `library(reticulate)` in my Rmarkdown file at all. It knits fine without it.) – Gil Henriques May 08 '21 at 16:49