- Notifications
You must be signed in to change notification settings - Fork1
msramalho/hackacity2019
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
The final presentation is availablehere
- Use regression techniques to uncover what makes local housing (Alojamento Local) locations be valued by customers and entrepreneurs so that the municipality can better understand how to approach the overflow of local housing problem and how to influence its evolution overtime
- Detect local housing establishments on booking.com that are not listed in the official municipality dataset, a helping hand in finding illegal establishments
- Implement interaction with Porto's APIs -01_api_interaction.ipynb
- These functions (and some more) have been extracted into an importable file -dataset_bending.py and there are tests and use cases available insome_tests.ipynb and the complete listing of datasets and sub-datasets available in the APIs is available atsub_datasets.ipynb
- Some useful functions to inspect and handle geographical data have also been implemented -geographical_functions.ipynb they have also been isolated to thegeographical_functions.py file
- Merge the local housing dataset with the data scrapped from booking.com -02_dataset_building.ipynb
- The 'scraping' of booking (with no financial goal) has been implemented inbooking_scrapping.py and some helper functions for this are present inbooking_parsing.ipynb
- Feature engineering from more geographical datasets into the main local housing dataset (green spaces, access to metro, bus, ...) -03_feature_engineering.ipynb
- Merge the feature engineering dataset with the local housing dataset -04_merge_with_features.ipynb
- Analysis of the built datasets, clustering, regression
A script to download all the datasets is available indownload_all.ipynb and the remaining data is available on thedatasets/ folder.
The temporary datasets that were built and used during the competition are available in thegenerated_datasets/ folder (Initially they were in the root folder and the notebooks still expect that).