Hi, I found a solution using *Wrapping*, adding a #+NAME to #+RESULTS block resolve the problem. This is not defined / described into the doc, and actually i found this process a little anti-intuitive. #+RESULTS has already a name, resulting/linked from/to the execution of mypythoncode ... Best regards #+NAME: attr_wrap #+BEGIN_SRC sh :var data="" :results output   echo "#+NAME: cities"   echo "$data" #+END_SRC #+NAME:mypythoncode #+begin_src python  :results value raw :output  :post attr_wrap(data=*this*) :return tabulate(df, headers=df.columns, tablefmt='orgtbl') import numpy as np import pandas as pd from tabulate import tabulate df = pd.DataFrame(np.random.randint(0,10,size=(10, 4)), columns=list('ABCD')) #+end_src Le 01/12/2021 à 09:04, Sébastien Rey-Coyrehourcq a écrit : > > Hi there, > > Perhaps it's a bug, or something i don't understand, i post on reddit > (https://www.reddit.com/r/emacs/comments/r5yt4a/r_talking_with_python_using_orgtable_not_work/) > to see if people on org community know the problem. > > I cross post here to see if it's a know bug or somethings ? > > There is something i don't understand, > a difference between behavior of org-babel and org-mode when you use > org table to pass data using or not using :var. > > |#+NAME:mypythoncode #+begin_src python :results value raw :output > :return tabulate(df, headers=df.columns, tablefmt='orgtbl') import > numpy as np import pandas as pd from tabulate import tabulate df = > pd.DataFrame(np.random.randint(0,10,size=(10, 4)), > columns=list('ABCD')) #+end_src #+RESULTS: mypythoncode | | A | B | C > | D | |---+---+---+---+---| | 0 | 0 | 9 | 6 | 0 | | 1 | 2 | 9 | 0 | 4 > | | 2 | 9 | 6 | 0 | 1 | | 3 | 6 | 1 | 8 | 1 | | 4 | 4 | 2 | 1 | 4 | | > 5 | 2 | 1 | 1 | 1 | | 6 | 4 | 8 | 9 | 0 | | 7 | 1 | 4 | 8 | 7 | | 8 | > 9 | 3 | 2 | 5 | | 9 | 5 | 0 | 7 | 3 | #+NAME:lib-R #+HEADER: :var > code=mypythoncode #+begin_src R :results output library(ggplot2) > library(dplyr) library(lubridate) str(code) #+end_src #+RESULTS: lib-R > : chr "| | A | B | C | D |\n|----+-----+-----+-----+-----|\n| 0 | 8 | > 0 | 5 | 2 |\n| 1 | 2 | "| __truncated__ | > > As you see, the org table is not recognized as a dataframe by R. > > If i replace by a basic org table : > > |||#+NAME: any_data | | parameter | value | |---+-----------+-------| | > 0 | heats | 30 | | 1 | heats | 30 | #+NAME:lib-R #+HEADER: :var > code=any_data #+begin_src R :results output library(ggplot2) > library(dplyr) library(lubridate) str(code) #+end_src #+RESULTS: lib-R > : 'data.frame': 2 obs. of 3 variables: : $ X : int 0 1 : $ parameter: > chr "heats" "heats" : $ value : int 30 30 ||| > > That works... > > Best regards, > > ||