By default, the Trend chevron is open. Here, you will see a chart preloaded with the dependent variable as the primary Y-axis (line) and the first incremental variable as the bar chart (secondary Y-axis).
Trend charts help you understand how KPIs and drivers evolve over time, aiding in shaping initial hypotheses and guiding variable selection for modeling.
Secondary Variables: By default, only one secondary variable is selected, but users can select up to 50 secondary variables. Ensure that selected secondary variables have similar units of measure.
Filter by: On the left side of the page, filter the data by specific dimensions (e.g., state level) to see trends for particular subsets of data.
Primary and Secondary Variables: Change the primary variable using the primary variable dropdown and select multiple secondary variables using the secondary variables dropdown.
Period Selector: Choose a subset of the entire data period for a more focused analysis. By default, it uses the entire duration of your data cube (e.g., January 2020 to December 2022).
Refresh Button: Click to load the updated chart after making selections.
Hover Functionality: Hover over the chart to see values for all variables for a specific date.
Legend Interaction: Click on variables in the legend to hide or show them on the chart. This is useful if some secondary variables have much higher values and overshadow others.
Below the chart, see the correlation between secondary variables and the primary variable. For example, if sales is your primary variable and you have selected 10 secondary variables, you will see the correlation of each secondary variable with sales. Changing the primary variable will update the correlations accordingly.
Download Data: Use the download icon to get the trend data.
Download Image: Click the three lines below the download option for a PNG image of the chart.
Save Trends: Save the current trend view by providing a name. Saved trends can be revisited and edited. The saved trends are accessible from the trend chevron, which shows all saved trends with details like the primary variable, secondary variables, and the selected period.
Access Saved Trends: Click on the Trend chevron to see all saved trends.
View Details: Hover over saved trends to see secondary variables and periods.
Download and Delete: Download data or delete saved trends using checkboxes and respective buttons.
Pagination: Use pagination at the bottom of the screen if there are many saved trends.
Filter Saved Trends: Use the filter option to search for trends containing specific variables or within a particular period
While trend charts show data at the most granular level (e.g., weekly), time comparison charts aggregate data for higher-level comparisons (e.g., yearly, quarterly).
Select Variables: Choose multiple variables for comparison.
Select Time Periods: Compare data across up to four time periods. Create custom periods if needed. For instance, compare yearly data by selecting different years or create a custom period by specifying start and end dates.
Refresh Chart: Load the chart by clicking the refresh button.
Slider Interaction: Adjust the slider to view data for a subset of variables. This is useful when dealing with a large number of variables.
Download Data: Use the download option to get time comparison data.
Below the chart, a data table shows the total values and percentage differences for each selected variable across the periods. This includes detailed comparisons between each period and the overall difference from the first to the last period.
Download Data: Use the download option to get time comparison data.
Save Time Comparisons: Save your time comparison view for future reference.
The review module, with its trend and time comparison features, allows for a comprehensive analysis of how variables evolve over time and how they compare across different periods. By utilizing these tools, users can develop informed hypotheses and refine their data modeling approach.
Event Peaks and Troughs: If unexplained peaks or troughs appear in the trend chart, check for corresponding events like holidays or other significant occurrences.
Split Variables: If there is a significant shift in data for a particular variable over different time periods, consider splitting the variable by period. This can help explain the model better and provide more accurate insights.
Advanced Settings: Remember that the advanced settings, including creating new variables, can be adjusted to better tailor the analysis to your specific needs.
Unit Consistency: Ensure that selected variables in trend charts have consistent units to facilitate meaningful comparisons.
This covers all functionalities within the review module, helping users effectively analyze and interpret their data.