Temitope Odeyemi

Know Your Charts, Tell The Right Story With It Understanding when and how to use the right visual as a data analyst. What is data visualization? This is the representation of data in various forms, from charts to graphs to maps and tables. These presentations allow anyone from a novice to an expert to identify trends and patterns in data. By having data visualization skills one would be able to effectively manage, understand, analyze, and communicate insights derived from dataset. Data visualization tools like Power BI, Tableau, Plotly, or Excel are great ways to tell stories, get insights, and provide solutions. However, do you really know the appropriate chart to use in different scenarios and contexts? Read along with me as I give different types of charts and the appropriate way they should be used. The first thing you need to know is that visualization is used for;     Comparison- To compare the magnitude of measures. Bar charts and line charts are helpful for comparison. Composition- This can also be known as part-to-whole relationship. This is used to determine the percentage variation in the data. This aims to show how individual parts of the datasets are combined to make up the whole dataset or a whole feature in the dataset. Treemap, pie charts, stacked bar charts, and stacked area charts are visuals to be considered here. Relationship- This is when the value of one feature could be used to predict the value of another feature. Scatterplots, bubble charts, and heat maps are best used here. Distribution- Distribution visualizations are used to show the spread of possible values for a variable. It shows how often these values occur. It can also be used to help identify outliers. Histogram, density plot, Box, and whisker plots are used to show distribution. Types of Visuals Bar chart - Bar charts are used to compare values for different categories, they are used in tracking changes over time and are best used when the changes are large. One can either use a horizontal or vertical bar chart, depending on the amount of text you want to display at the foot of each bar. Line charts - these are used in tracking changes over a period of time for one or more groups (each group would have its own line) and are best used when the changes are small. They are basically used for time series data. Line charts are ideal for displaying continuous data. Stacked bar chart- Like bar charts, it compares different categories. Each bars are then divided into subcategories. This is used to compare totals across categories and also to show how each subcategory forms a part-to-whole. Stacked area chart- This is an extension of a line chart. It is used to compare categories over time, showing the composition of each category. The part-to-whole would be clearly indicated by filling the sections between the lines and the x-axis with different colors. Pie charts- This is the most common type of composition chart. It is used to show basic composition. However, it is limited by how many parts one can display. When categories are close in size, it is quite difficult to tell which is bigger and it becomes hard to read. If it would divided to more than 4 parts, then a pie chart should not be selected. Tree map. Just like a pie chart, it is used to show composition. Instead of a circle, a rectangle is used. When a whole consist of 5 or more parts, a treemap is encouraged. The rectangles would be ordered in decreasing size from the top left to bottom right. It can also be a multi-level visualization chart as each rectangle or category can further be divided into sub categories. Scatterplot- This is used to show relationships between only 2 features. The values would be plotted at single points. The x and y values of each point on the plot responds to the values of the 2 features to be considered. If we have a dataset of 120 rows, we would have 120 points. Bubble chart- For 3 relationship features, a bubble chart should be selected. With 2 features plotted the same way as a scatter plot and the 3rd feature would determine each plotted point size, creating bubbles of various sizes. Heat map- A heat map is used to determine which 2 features are related to each other in a time efficient manner. A correlation value between features are used. Correlation is the statistic that measures how strongly or weakly 2 features are related. It is measured from -1 to 1, with 0 representing no relationship. These values are then plotted as color blocks on a square grid. The darker the color, the stronger the relationship. Histogram- This looks very much like a bar chart, but it groups numbers into intervals. The height of each bar show how many data points falls into each intervals. There are no spaces between bars in histograms. It is used when the data are numerical and you want to see the shapes of the dataset distribution. Tables- Tables allow for a detailed data view in a structured format, presenting information in rows and columns similar to spreadsheets. They are essential for displaying numerous variables at once or when specific figures are required. Cards- Cards display a single value in a large font and are often used to highlight a key figure, like total sales or average cost. They are less detailed but excellent for drawing attention to priority data. Column charts- These are perfect for comparing values across categories shown as bars. They are best used with multiple data series and where highlighting relative proportions is necessary. It is used to show comparison. Funnels- funnels help visualize a process that has stages. A funnel chart is shaped like a funnel with the first stage being the largest, and each subsequent stage smaller than its predecessor. Gauge- A gauge chart has a circular arc and displays a single value that measures progress toward a goal. The goal, or target value, is represented by the line (needle). Progress toward that goal is represented by the shading. And the value that represents that progress is shown in bold inside the arc. All possible values are spread evenly along the arc, from the minimum (left-most value) to the maximum (right-most value). KPI- A Key Performance Indicator (KPI) is a visual cue that communicates the amount of progress made toward a measurable goal. Slicer- A slicer is a standalone chart that can be used to filter the other visuals on the page. You can reach out to me on LinkedIn at Temitope Odeyemi or Twitter (here) Check out my GitHub page here to see several projects that I have done. As it is commonly said in Italy, Ciao for now 😊