Repository for practice simple pipeline for wrangling and preprocessing data

Autos data wrangling and preprocessing

The aim of this project is to create simple pipeline for data wrangling and preprocessing autos dataset. So, in the future when I work with this autos dataset again, the data wrangling and preprocessing process automatically follow this pipeline step or I can develop for the another purpose

Process

Here below are the step of process to create pipeline for data wrangling and preprocessing autos dataset :

• Install package requirements for this project
• Read csv file into dataframe
• Rename some columns name into snake case type
• Change value type into datetime format for columns (ad_created, date_crawled, last_seen)
• Perform the cleaning data for column price and odometer, so the value from that column we can change into integer format
• Drop the column seller and offer_typ because the unique value of that column are inbalance
• Drop the column num_of_pictures because all the value is 0
• Drop the column name and postal_code because that columns less useful for our purpose
• Handle outliers for price columns
• Handle missing value. For column that has value type an object, fill missing value with mode. For column that has value type a numeric, fill missing value with median
• Perform the reset index
• perform the normalization process for columns with numeric data type, except for the price column because we will make it the target column in the future
• Perform the encoding process for columns of categorical data type
• Save the data that has been cleaned and pre-processed into csv format

Review and Feedback

I really appreciate for the review and feedback. Let's connect with my linkedIn account.