teal.data with Python

Overview

It is possible to use python code (via the reticulate package) to create a teal dataset. Therefore if a python package is required to access/create your data, it can still be used in a teal application.

We recommend having a thorough understanding of the reticulate package (see here) before using the python functionality described below.

Example

In this section we show a simple example of how python code can be used to create a teal dataset.

For this example we require a python environment with pandas installed. See here for further details.

We first load the packages:

library(reticulate) # may need to be installed if not available
# reticulate function to call to install the pandas package if not installed: py_install("pandas")
library(teal.data)

Next we define the python code which we want to use to create the dataset. In this example, for demonstration purposes we create a trivial dataset which could easily be done directly in R. Note we use a variable num_rows which needs to be passed to the python code from R.

python_code <- "import pandas as pd
data = pd.DataFrame({\"id\" : range(r.num_rows), \"val\" : range(r.num_rows)})"

Next we create a python_dataset_connector object.

x <- python_dataset_connector(
  dataname = "DATA", # the teal dataset name
  code = python_code, # the code used to generate the dataset
  object = "data", # the object in the python code to be converted to a data.frame for the teal dataset
  keys = "id", # the key for teal dataset object
  vars = list(num_rows = 5L) # any variables passed from R into python (note this could be an R variable)
)

Finally we can test the code by pulling the data into R:

x$pull()
print(x)

Further Concerns