WEB APPLICATION DEPLOYED AND IN PRODUCTION live at url https://housing-prices-predictor.onrender.com/
- because API can both be used as an application with a working UI and as an endpoint navigating instead to url https://housing-prices-predictor.vercel.app/predict/json with the necessary data payload using postman will return a json response. But navigating to https://housing-prices-predictor.vercel.app/predict once input data is entered will redirect to the base url https://housing-prices-predictor.vercel.app/ with the predicted value
- local machine usage will be to simply navigate to
/server-side
directory then assuming all dependencies are installed by following Source Code Usage instructions below just run python index.py and wait forlocalhost/127.0.0.1
server at port5000
to start e.g.https://127.0.0.1:5000
- clone repository with
git clone https://github.com/08Aristodemus24/housing-prices-predictor.git
- navigate to directory with
readme.md
andrequirements.txt
file - run command;
conda create -n <name of env e.g. housing-prices-predictor> python=3.11.2
. Note that 3.11.2 must be the python version otherwise packages to be installed would not be compatible with a different python version - once environment is created activate it by running command
conda activate
- then run
conda activate housing-prices-predictor
- check if pip is installed by running
conda list -e
and checking list - if it is there then move to step 8, if not then install
pip
by typingconda install pip
- if
pip
exists or install is done runpip install -r requirements.txt
in the directory you are currently in
- instead of a linear model implement a function to engineer new features that results in a more polynomial equation to use as our model:
- note that we have to normalize data first before passing data to map_feature() which engineers new features out of the current features to make the equation more polynomial
- I'm actually passing unnormalized X values in the test model so I need to find a way to recover previous standard deviation and mean calculated from the training the data which was used to normalized both training and cross validation data