Data Analytics Internship - Week 1 & 2 Report
- Cleaning Challenges Encountered E-commerce Dataset: - Missing Values: CustomerID had 135,080 missing values (filled with 0), Description had 1,456 missing values (rows removed). - Invalid Values: 9,725 rows had negative/zero quantities, and 1,058 rows had negative/zero prices (removed). - Duplicates: 5,268 duplicate rows were removed. - Date Format: InvoiceDate was converted to datetime format. Netflix Dataset: - Missing Values: Director (2,634), Cast (825), Country (831), Rating (8), Date_Added (10) were filled with 'Unknown' or mode. - Duplicates: No duplicate rows found. - Text Standardization: Column names converted to lowercase, Country to uppercase, Title to title case. - Date Format: Date_Added converted to datetime format.
- Key EDA Findings E-commerce Dataset: - Total Revenue:$8.9M+ - Best-Selling Product:"WHITE HANGING HEART T-LIGHT HOLDER" (231,000+ units sold). - Top Country: United Kingdom generates the highest revenue. - Seasonal Trend: Sales peak in November and December. Netflix Dataset: - Content Distribution: 69.6% Movies, 30.4% TV Shows. - Peak Year for Content Addition: 2019 (2,016 titles added). - Top Country: United States produces the most content. - Most Common Rating: TV-MA.
- Most Common Genre: Dramas.
- Top Insights E-commerce Insights:
- United Kingdom Dominates Revenue: The UK is the primary market, generating the highest revenue.
- Seasonal Sales Peaks in November-December: Holiday season drives higher sales.
- Small Quantities per Transaction: Most customers buy in small quantities (1-10 units). Netflix Insights:
- Movies Dominate Netflix Content: 69.6% of content is Movies.
- Content Addition Peaked in 2019: Netflix aggressively expanded content in 2019.
- Focus on Modern Content: Most content was released between 2010-2020.
- Conclusion: Both datasets were successfully cleaned and analyzed. The E-commerce dataset revealed key sales trends and customer behavior, while the Netflix dataset provided insights into content distribution and production trends. The cleaning process addressed missing values, duplicates, and invalid entries, ensuring data quality for analysis.