Hospitalization Prediction Model Improvement

Apply multiple methods taught in Advanced Python(DS3001) to speed up original project code.

The Python files

There 5 python files:
01_Merged_original_Step1.ipynb
02_Merged_original_Step2.ipynb
03_Merged_improved_Step1.ipynb
04_Merged_improved_Step2.ipynb
05_Seperated_original&improved_Step1&Step2.ipynb

How to run the files:

Step1, run 01 and 02 ipynb files - Original running time

01 and 02 ipynb files return a total running time of original codes.

Step2, run 03 and 04 ipynb files - Improved running time

03 and 04 files return a total best running time of improved codes.

Step3(Optional), run 05 ipynb files - Detailed comparing & alternative methods

05 file provides a detailed comparing of between original codes and improved codes, chunks by chunks. Besides the best performance methods, this file also provides alternative methods which are heuristic, e.g comparing the performance between differenct Cython versions and multiprocess.

You need to put the ipynb files and dataset under the same directory, otherwise you need to specify to loading path of dataset in the ipynb files before you run them.

Dataset

You can find the dataset in the following links: https://drive.google.com/open?id=1aJs25Fs2Nd_sGwew-B03tOVa2n9kpdxb

In the Dataset folder, there are 7 csv files, please download all of them.

Requirements

Python 3.6

Python packages:

• numpy
• pandas

Notes:

The run time reported in the report might slightly differ from the time shown on ipython notebook due to the computer or time it was last ran. Please note that the run time you get might also slightly differ due to performance of different device.