Candies for Halloween Party

For my Maven Halloween Challenge project, I employed Power BI to analyze and visualize data from 85 types of candy, helping determine the top 3 treats to hand out during Halloween. Here’s a breakdown of my technical steps: For this project, I used Power BI to analyze and visualize data from 85 types of candy, to find top 3 treats to distribute during Halloween. Below mentioned are the steps I followed:

1. Data Preparation and Visualization Setup:
   • Loaded the pre-cleaned dataset into Power BI, used Power Query Editor to apply rounding and percentage transformations on the key columns: sugar percentage, price percentage, and win percentage. Doing this gave a clarity as the values given were like this 0.00675903 so had to apply some changes
   • Created three cards to display the totals of the major candy types: 21 bars, 37 chocolates, and 14 caramel candies.
2. Price, Sugar, and Win Percentage Comparison:
   • For price analysis, I used a line chart to compare the sum of price percentage against candy competitors, applying a Top N filter to focus on the top 3 by price percentage.
   • For sugar content, I used a bar chart to compare the sum of sugar percentage by competitor, again filtered to show the top 3 by sugar percentage.
   • A pie chart was employed to visually represent the win percentage, highlighting the top 3 performing candies based on this metric.
3. Scatter Plot for Multivariate Analysis:
   • I used a scatter plot where:
   • X-axis represents the minimum of price,
   • Y-axis represents the minimum of sugar, and
   • Legend shows the maximum win percentage. This was crucial in understanding the relationship between price, sugar, and win percentages for the different competitors.
4. Candy Type Breakdown: I found certain columns (e.g., bar, caramel, fruity) to be of limited use and therefore filtered them out to focus on the most relevant data points. The visualization clearly highlights the top 3 candies based on price, sugar, and win percentages, helping guide decisions for Halloween candy distribution.
   • I now plan to create a grouped bar chart that breaks down the top 3 candies based on their candy types (e.g., chocolate, caramel, bar) to provide a detailed profile of their characteristics.
   • Additionally, I intend to use a treemap to visually display the minimum sugar by competitor name, offering a clearer view of how each candy performs in terms of sugar content.