# rainfall-prediction

**Repository Path**: shiji203/rainfall-prediction

## Basic Information

- **Project Name**: rainfall-prediction
- **Description**: Rainfall prediction models (Linear and Logistic) trained on publicly available datasets from Austin, Texas
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2020-06-21
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

This repository builds a Linear as well as a Logistic model to predict rainfalls in Austin, Texas.

The following dataset constitutes 3.5 years worth of weather data, including temperature, humidity, dewpoints, etc: https://www.kaggle.com/grubenm/austin-weather

Execute the files **linearRegression.py** and **logisticRegression.py** to obtain predictions for an arbitrary day with hardcoded input parameters.

![5b3f34740ce87](https://i.loli.net/2018/07/06/5b3f34740ce87.png)

A day (in red) having a precipitation of about 2 inches is tracked across multiple parameters.

![5b3f34511cde8](https://i.loli.net/2018/07/06/5b3f34511cde8.png)

Manually classifying the precipitation levels into 4 different classes as follows:

- No Rain: precipitation<0.001

- Drizzle: 0.001<=precipitation<0.1
- Moderate Rains: 0.1<=precipitation< 1.2
- Heavy Rains: precipitation>=1.2

The graphs we obtain after classifying express various trends which tie rainfall and humidity, visibility and temperature together.

![5b3f36bd69c75](https://i.loli.net/2018/07/06/5b3f36bd69c75.png)