Overview / Usage

Do you load, process, organise, and analyse massive amounts of data with Pandas? Try running it 10, 20, and 30 times. and occasionally quicker? No coding modifications are necessary; Intel Modin, a component of the oneAPI AI Analytics toolset, will assist you in achieving that. The utilisation of Modin, how to obtain it, and how it can speed up your code are all covered in this brief project, along with some real-world applications and performance comparisons.

More details: https://devmesh.intel.com/post/1034061/how-intel-oneapi-ai-analytics-toolkit-ai-kit-accelerates-ai-pipelines

Methodology / Approach

Modin can be obtained in a variety of ways. Logging into the Intel DevCloud might be the simplest option. Get access in just a few clicks, and once you click the email and open one of the samples, everything is installed. You may immediately put it to use. You may also look for it on GitHub by searching for "Modin Git" in the search bar. I decided to set up the oneAPI AI-toolkit. I'm using an Intel Xeon computer. I will compare Pandas and Modin's performance using a real-world Kaggle dataset.

Technologies Used

Numpy, Scikit-learn, XGBoost, Modin, oneAPI

Repository

https://www.kaggle.com/code/rishirajacharya/oneapi-modin-gives-pandas-wings

Other links

• AI Kit