This project contains codebase for data preparation and summerization before passing it to the machine learning model for classifcaion.

1. Python
2. Pandas
3. Numpy
4. tqdm
5. sci-kit learn

• open your terminal & cd into your project directory
• run git clone https://github.com/Databreedteam/all-state-project.git or download the zip from this location https://github.com/Databreedteam/all-state-project
• cd into the all-state-project cloned directory
• run pip install -r requirements.txt from your terminal to install the dependencies
• finally, run python data_wrangler.py and wait for the progress bar to finish
• go to the result folder, you will see the computed summary of the train csv file.

To initialize this class, use the code below

Data = DataWrangler()

This method should be called immediately after Initializing the class.

full_df, quote_df, purchased_df = Data.convertToDataframe()

It accepts the train.csv as input then return a dataframe as output.

full_df , quote_df, purchase_df = Data.splitDataFrameToSections()

This method recive dataframe as input then return 3 different dataframes types as output.. which are :

1. full_df variable contains Full dataframe containing all quotes and purchased insurance.
2. quote_df variable contains dataframe of all quotes insurance only.
3. purchased_df variable contains dataframe of all purchased insurance only.

df = Data.prepareDataForOperation(quote_df)

This method accepts dataframe and performs operation such as filling or removing missing variables on the colums and return the updated dataframe.

df = Data.transformstateToCensusRegion(df)

This method takes in a datafram then convert the state colum into US censorship region by grouping them into (mid_west, north_east, south, west) then return an updated dataframe.

df = Data.transformAgeToCategorical(df)

This methods takes in a dataframe the transform the age from numerical data to categorical data by diving them into (LOW, MED, HIGH) then return the updated dataframe.

df = Data.performWeightedAverageOnAG(df)

This method takes in a dataframe and compute the weighted average on all the insurance policy features.. i.e colum A-G, then return the updated dataframe

df = Data.transformTimeOfTheDay(df)

This method takes in a dataframe then transform the time of the quote to a categorical values (MORNING, AFTERNOON, NIGHT) then return updated dataframe.

df = Data.performWeightedAverageOntTime(df)

This methods takes in a dataframe and computes the weighted average on the time of the quotes and return an updated dataframe.

df = Data.summerizeQuote(df)

This method takes in a dataframe and summerize duplicates entry quotes for customers with same customer_ID then return the updated dataframe.

It will also drop some colum that wont be needed.

Data.saveToDirectory(df, "summerized_result.csv")

This method can be called at anypoint in your implementation of this code to save the dataframe.

It takes in a dataframe and also a filename ending with a .csv extension and save the result of the dataframe as csv file into the result folder.