DonaldRR/housePricePred

housePricePred

Code for House Prices: Advanced Regression Techniques(https://www.kaggle.com/c/house-prices-advanced-regression-techniques)

Environment & Dependencies:

• Python: 3.6
• numpy(1.14.5)
• pandas(0.23.4)
• scipy(1.1.0)
• matplotlib(2.2.3)
• xgboost(0.80)
• seaborn
• sklearn(>=0.19.2)

(Versions are not limited as stated above, they are just my laptop's configuration. Later versions are likely to work.)

Install

Install project in your local computer. (Install git before, with pip install git)

git clone https://github.com/DonaldRR/housePricePred.git

Datasets

You need to download Datasets from kaggle's House Prices:Advanced Regression website, or using command line (if this works...)

kaggle competitions download -c house-prices-advanced-regression-techniques

By default, you need to make two directories(datasets and processed_data) under your project directory CurrentProjectDir to store data, like

-->CurrentProjectDir
  -->datasets
  -->processed_data

Of course, you can change directories of datasets in config.py

Run Scripts

Preprocess

First, process your original CSVs.

python preprocess.py

Processed files will occur under processed_data directory

Run Models

Run specific model to train the data. Arguments are available,

model_type, 'nn' or 'xgb' available

python run_model.py [model_type [model_type [model_type [...]]]

More

TODOs in model.py are what you need to complete.