Tensorflow Tutorial in RStudio Classify Structured Data with feature columns

Question

I am working to learn TensorFlow for Machine learning.

I am using R 4.1.2, RStudio 2022.02.0-443, Python 3.9.7 on Ubuntu 21.

I have installed all the libraries on the Rstudio side following all the installation instructions from https://tensorflow.rstudio.com/installation

On the virtual environment side I cannot find the equivalent of libraries "rsample" and "Matrics". I have installed matrix by pip install matrix.

I have updated all the packages.

Next I went to https://tensorflow.rstudio.com/tutorials/ and started going through tutorials.

This is to ensure that my installation is working and I understand what libraries are needed to do what before getting into what I want to use this for.

So far:

1. The Overview tutorial with MNIST dataset worked fine.
2. The Basic Regression with Boston Housing Prices dataset worked fine.
3. Next I tried to learn the tutorial as it is very close to what I will be using this software for my 1st model and application. https://tensorflow.rstudio.com/tutorials/advanced/structured/classify/

The data is not available from the source as it says in the tutorial. So I got the same data from Heart Data csv file

Up to the following line my output looks the same as in the tutorial.

str(spec_prep$dense_features())

I get the following error when I execute the above line:

Error in py_call_impl(callable, dots$args, dots$keywords) : 
ValueError: column_name: thal vocabulary dtype must be string or integer. dtype: <dtype: 'float64'>.

Something is happening in the variable manipulation in the lines above. I have no clue how to fix this, as I have not written any code and upto this point output agrees with the tutorial.

Any suggestion?