- Preliminary Data Cleaning of 'kz.csv' Dataset
This project involves the cleaning and preprocessing of a dataset ('kz.csv') containing information related to orders. The primary goal is to handle missing values, convert data types where necessary, and ensure consistency in certain columns, such as 'category_id,' 'price,' 'user_id,' 'category_code,' and 'brand.'
The stakeholder(s) for this project are those interested in analyzing order data. The business problem addressed is ensuring the quality and reliability of the dataset for subsequent analyses. No external research citations are included.
The dataset ('kz.csv') contains information on orders, and columns such as 'category_id,' 'price,' 'user_id,' 'category_code,' and 'brand' are present. Descriptive statistics provide insights into numerical features, and missing values are addressed in 'category_id,' 'price,' and 'user_id.' Numeric data stored as strings in 'category_code' and 'brand' are handled, and rates of invalid entries are calculated.
The following steps were taken to complete the project:
- Importing necessary libraries (Pandas, NumPy, Datetime).
- Loading and displaying the dataset.
- Calculating descriptive statistics.
- Displaying data types of each column.
- Handling missing values in 'category_id,' 'price,' and 'user_id.'
- Filling missing values based on relevant information.
- Processing 'category_code' and 'brand' columns for consistency.
- Calculating rates of invalid entries in 'category_code' and 'brand.'
- The project assumes default values based on the dataset context.
- Functions for handling numeric data stored as strings can be adapted for similar scenarios.
- Questions or improvements are welcome, feel free to reach out.
This project is part of a portfolio for a data analyst job search, demonstrating skills in data cleaning and preprocessing.