Before going through this article, you should have some basic knowledge of R programming, Python programming(Pandas), R Studio and R markdown.
| There always seems to be a perennial debate on which programming language is better. Both languages, R and Python, have their own communities and support groups. None of them are wrong. R and Python both have their own pros and cons and each language is better suited for different scenarios. Rather than debating on which language is better, in my opinion, we should try to incorporate both languages in our work. That is where “Reticulate” package becomes handy. Similar kinds of libraries(rpy2) are available in Python to incorporate R in Python and in R, other packages are also available. |
Let’s get started. You need to install Reticulate package Installation and setup and install Pandas Installing python packages first.
The first step is loading your library(reticulate) and setting path to a ‘conda’ executable. In this case, appropriate ‘conda’ is automatically set. To know more, execute ?reticulate and ?conda_python in your R console.
Note: Rmarkdown can execute both R and Python codes. R codes are enclosed in {r} chunk and python codes are enclosed in {python} chunk.
#R code
library(reticulate)
conda_python()
[1] "C:\\Users\\Sanjay\\AppData\\Local\\r-miniconda\\envs\\r-reticulate\\python.exe"Lets load a csv file cars.csv . I already have a “cars.csv” file in the given path below. You can also load any csv file you want.
#Python code
import pandas as pd
car_python = pd.read_csv("cars.csv")
car_python.head(n=10)Before going through this article, you should have some basic knowledge of R programming, Python programming(Pandas), R Studio and R markdown.
There is always a perennial debate on which programming language is better. Both languages, R and Python, have their own communities and support groups. None of them are wrong. Both R and Python languages have their own pros and cons and each language is better suited for different scenarios. Rather than debating on which language is better, in my opinion, we should try to incorporate both languages in our work. And Reticulate package in R helps to incorporate Python codes in R. Similar kinds of libraries(rpy2) are available in Python to incorporate R codes in Python.
The goal of this article is to get you started with the basic interaction between R and Python dataframes using Rmarkdown and Reticulate package. In this article, we are going to cover the following things.
- What is reticulate?
- Installation and setup for this demo
- Loading library and setting conda
- Using python code to import a csv file
- Adding a new column in dataframe
- Source Python code or File
- Convert dataframe in R to Python dataframe
- Interactive graphs in R with python dataframe
What is reticulate?
Reticulate is an R package that can assist us to interact between R and Python. It is a bridge between R and Python.
Installation and setup for this demo
Let’s get started. Firstly, you need to install Reticulate package (Installation and setup)and then install Pandas (Installing python packages) for this short demo.
In your console, execute:
install.packages(“reticulate”) and then
reticulate::conda_install(“r-reticulate”, “pandas”)
Rmarkdown can execute both R and Python codes. You can write both R and Python codes in their respective code chunks:
Loading library and setting conda.
Lets load the library(reticulate) and set the path to a ‘conda’ executable. In this case, appropriate ‘conda’ is automatically set. To know more, execute ?reticulate and ?conda_python in your R console.
#R code
library(reticulate) #<<
conda_python() #<<[1] "C:\\Users\\Sanjay\\AppData\\Local\\r-miniconda\\envs\\r-reticulate\\python.exe"It by default sets the version of Python found on your path. But you can also set the path to your desired python version.
Using python code to import a csv file.
Copy the path in the link cars.csv to provide path to your csv file.
#Python code
import pandas as pd
car_python = pd.read_csv("cars.csv")
car_python.head(n=10) speed dist
0 4 2
1 4 10
2 7 4
3 7 22
4 8 16
5 9 10
6 10 18
7 10 26
8 10 34
9 11 17<<<<<<< HEAD
Adding a new column “speed_categories” using python in “car_python” dataframe to give categories to speed.
#Python code
car_python["Speed_category"]= ""
for i in range(len(car_python)):
if car_python.loc[i, "speed"]>0 and car_python.loc[i,"speed"]<=10:
car_python.loc[i, "Speed_category"]= "Low speed"
elif car_python.loc[i, "speed"]>10 and car_python.loc[i, "speed"]<20:
car_python.loc[i, "Speed_category"]= "Medium speed"
else:
car_python.loc[i, "Speed_category"]= "High speed"
car_python.head(n=10)Adding a new column in dataframe.
Let’s add a new column->“Speed_category” in our dataframe using pandas. - If the speed is less than 10 and greater than zero, label it as “Low speed”. - If the speed is between 10 and 20, label it as “Medium speed” - Label the rest as “High speed”. We will use this column later to create a 3D graph.
#Python code
car_python["Speed_category"]= ""
for i in range(len(car_python)):
if car_python.loc[i, "speed"]>0 and car_python.loc[i,"speed"]<=10:
car_python.loc[i, "Speed_category"]= "Low speed"
elif car_python.loc[i, "speed"]>10 and car_python.loc[i, "speed"]<20:
car_python.loc[i, "Speed_category"]= "Medium speed"
else:
car_python.loc[i, "Speed_category"]= "High speed"
car_python.head(n=10) speed dist Speed_category
0 4 2 Low speed
1 4 10 Low speed
2 7 4 Low speed
3 7 22 Low speed
4 8 16 Low speed
5 9 10 Low speed
6 10 18 Low speed
7 10 26 Low speed
8 10 34 Low speed
9 11 17 Medium speed<<<<<<< HEAD
How to convert dataframe in python to dataframe in R? Use function: “py$name_of_dataframe”
#R code
R_cars<- py$car_python
head(R_cars, n=4)
Converting dataframes in python to R
Use py$:
Here, car_python is a variable in python that stores a dataframe, which in the code below, is used by R and stored as a dataframe in an R variable, R_cars.
#R code
R_cars<- py$car_python
head(R_cars, n=4) speed dist Speed_category
1 4 2 Low speed
2 4 10 Low speed
3 7 4 Low speed
4 7 22 Low speed<<<<<<< HEAD
How to source Python code and use it in R?
I have a file time.py, which contains the following python function to convert speed and distance to time. In the next step, I’ll show how to use the time() function from “time.py” file to our R code :
# time.py
def time(DIST, SPEED):
return DIST/SPEEDYou can source python files using source_python() and use the functions from python files directly to your R code.
#R code
source_python("time.py")
for(i in seq(nrow(R_cars))){
R_cars[i, "Time"]=
time(R_cars[i, "dist"], R_cars[i, "speed"])
}
head(R_cars)
Source Python code or File
A python file, time.py contains the following python function to convert speed and distance to time. Download it. In the next step, we will see how to use the time(dist, speed) function from “time.py” file to our R code :
# time.py
def time(DIST, SPEED):
return DIST/SPEEDUse source_python(“Path to your python file”) and use the python functions directly to your R code.
#R code
source_python("time.py")
for(i in seq(nrow(R_cars))){
R_cars[i, "Time"]=
time(R_cars[i, "dist"], R_cars[i, "speed"])
}
head(R_cars) speed dist Speed_category Time
1 4 2 Low speed 0.5000000
2 4 10 Low speed 2.5000000
3 7 4 Low speed 0.5714286
4 7 22 Low speed 3.1428571
5 8 16 Low speed 2.0000000
6 9 10 Low speed 1.1111111How to convert dataframe in R to dataframe in Python? You can use “r.name_of_dataframe” function.
#Python code
python_cars2 = r.R_cars
python_cars2.head(n=5)Convert dataframe in R to Python dataframe.
Use ‘r.’ .
In the code below, R_cars is a variable in R which is called in a python code and converted to python dataframe and stored in a variable, python_cars2.
#Python code
python_cars2 = r.R_cars
python_cars2.head(n=5) speed dist Speed_category Time
0 4.0 2.0 Low speed 0.500000
1 4.0 10.0 Low speed 2.500000
2 7.0 4.0 Low speed 0.571429
3 7.0 22.0 Low speed 3.142857
4 8.0 16.0 Low speed 2.000000Using python data frame to create graph in R using ggplot2 and ggplotly.
Using python data frame to create 3D graphs in R.
# R code
my_plot3<-plot_ly(data = py$python_cars2,
x=~speed,
y=~dist,
z=~Time) %>%
add_markers(color= ~Speed_category)
my_plot3
To learn more about Reticulate package, visit Reticulate.
Interactive graph in R with python dataframe
Here, we are creating a scatter plot, line plot and a regression plot combined using a python dataframe,“python_cars2” in R code. -ggplot creates visually appealing graphs in R and ggplotly() function converts ggplot graphs to interactive ones.
#R code
library(ggplot2)
library(plotly)
my_plt<- ggplot(data = py$python_cars2,
aes(speed, dist))+
geom_point(aes( color= Speed_category, shape= Speed_category))+
geom_line(color= 'orange', alpha=0.5, linetype='dashed')+
geom_smooth(se= F, color='red')
my_plt %>% ggplotly()3D plots using plotly and python dataframe.
# R code
my_plot3<-plot_ly(data = py$python_cars2,
x=~speed,
y=~dist,
z=~Time) %>%
add_markers(color= ~Speed_category)
my_plot3To recapitulate, we learned about Reticulate and also viewed a short demonstration about interoperability between R and Python using the Reticulate package.
>>>>>>> ef957d37684dcbb30421b3805763f14e82e10b6eTo learn more about Reticulate package, visit Reticulate.
Corrections
If you see mistakes or want to suggest changes, please create an issue on the source repository.
Citation
For attribution, please cite this work as
Hamal (2022, March 7). Sanjay Hamal: Not just 'R or Python' but 'R and Python'!. Retrieved from https://sanjayhamal.com.np/post2022/2022-03-07-aboutreticulate/
BibTeX citation
@misc{hamal2022not,
author = {Hamal, Sanjay},
title = {Sanjay Hamal: Not just 'R or Python' but 'R and Python'!},
url = {https://sanjayhamal.com.np/post2022/2022-03-07-aboutreticulate/},
year = {2022}
}