Closed Aratiganesh123 closed 4 months ago
The only way to do that is to write the C++ DataFrame into a file in csv2
format which is the same format that Pandas can also read. Alternatively, you can do the whole thing in C++ and skip all the language conversions.
I am working on a project to speed up data loading and preprocessing using a C++ Dataframe library. My goal is to preprocess the data in C++ and then use the processed data to train models in scikit-learn and PyTorch. I am considering using Pybind11 to integrate C++ and Python.
My main concern is the overhead of converting the C++ Dataframe to a Pandas DataFrame in Python. I want to ensure that this conversion process is efficient and does not introduce significant overhead. Could you provide guidance or suggestions on how to achieve this with minimal overhead?