This Jupyter notebook is a visual data analysis investigating a handful of correlations across product categories in this dataset of 230 million Amazon reviews.
The following are a series of draft notebook I used for EDA (exploratory data analysis). Many pieces of the functions in the above notebook were drafted in these.
- pandas_feather_benchmark - a progressively scaled benchmarking comparison of
feather-formatVSpd.read_csvand.applyVS pandasdatetime - Time_Analysis - a time based analysis of several product catagories to discover any trends across annual, monthly, or weekly reviews
- Review_Timing - comparing across a series of time increments (day of week, day of month, and monthly) the average rating
- Word_Count - comparing common language in reviews across several catagories, included an analysis of top words