BigConnect Discovery Basics
This tutorial will teach you the basics of data preparation and visualisation with BigConnect Discovery. The main points addressed below are:
  • Importing and Preparing a dataset
  • Creating dashboards and visualisations

Preparing data

1.Login to BigConnect Discovery by going to http://localhost:8180 and use admin/admin for username and password. You will be taken to the Admin Workspace.
2.Open the menu and choose MANAGEMENT / Data Preparation / Datasets. On this page, click on the Generate new dataset link.
3.Choose My file to upload your data
4. Download the sample sales.csv dataset from https://github.com/bigconnect/demos/raw/master/discovery/sales.csv , upload it and choose Next. Leave the other options as default.
5. Review the dataset and choose Next to continue.
6. On the next screen, accept the defaults and click Done.
7. After import, a new Data Preparation Flow will be created and you will be taken to the flow screen.
8. Let's do some changes to the dataset:
8.1 Change the date format for OrderDate to yyyy-MM-dd. Click the three dots icon next to the OrderDate column and choose Alter / Set Format:
8.2 Set yyyy-MM-dd as the format in the bottom pane and click Add.
8.3 Delete the Country, orderprofitable and SalesAboveTarget columns. For each one of the, click on the the three dots icon next to the column name and choose Drop.
8.4 Set the date format for ShipDate to yyyy-MM-dd. Follow the same instructions in step 8.1 to do so.
8.5 Combine the latitude and longitude columns in a new column of type Point. The Point datatype expects data to be in WKT format, eg. POINT(lat lon). Without selecting a colum, choose derive from the Add rule section in the bottom pane:
1
concat('POINT (', `longitude`, ' ', `latitude`, ')')
Copied!
Enter Geolocation as the new column name and click Add.
The new column is added to the dataset:
9. Let's generate a new dataset based on the rules we set. Click on the Snapshot button in the top pane.
10. Leave all options to their defaults and click Done.
11. The snapshot will be created and will be displayed in the SNAPSHOT section of the Flow.
12. Click on the created snapshot to see rule execution results:
Everything went smooth, we don't have any errors.
13. Now let's create a Datasource from this snapshot. Click on the Create datasource button and you will be taken to the Datasource Schema configuration screen.
14. BigConnect Discovery makes some guesses as to how the columns will be mapped, but we need to review them. Let's make the following changes:
14.1 Click on the PostalCode column and change its Role to Dimension and Type to String.
14.2 Click on the latitude column and change its Role to Dimension and Type to Latitude
14.3 Click on the longitude column and change its Role to Dimension and Type to Longitude
14.4 Click on the Geolocation column and chang its Type to Point.
14.3 Make sure that OrderDate is selected as the Time-Type column and hit Next.
15. On the next screen, choose Day for Query Granularity and Month for Segment Granularity. Choose false for Rollup if enabled and click Next.
16. On the next screen, enter salesdata as the Datasource Name and hit Done.
17. The system will start to ingest the data in the Druid engine and the progress is reported. After finishing, you will see something like this:

Visualizing data

Now that we have ingested the prepared data into our Druid engine, it's time to create some dashboards and visualizations.
1.Open the menu and choose WORKSPACE / Admin Workspace to go to our personal workspace. On this screen click on Workbook.
2.Give a name to our new Workbook (eg. Demo) and click on Done.
3. Now let's add our newly created datasource to the Workbook. Click on the Add datasource baloon, choose salesdata and hit Done.
4. Click Next to move to the next screen. Enter Sales Dashboard as the Name for our new dashboard and click Done to finally open our new Workbook and Sales Dashboard.
5. Now let's add some charts. Click on the Chart dropdown button and select Chart.
6. Choose Pie Chart. Click on the Sub_Category Dimension and Sales Measure from the left pane to add them to our chart.
7. Click on the color palette and choose the blue color scheme for our pie chart:
8. Now let's add some labels to our chart. Click on the Data Label settings option, enable Show Data Label, select Series Name and Series Percent and enable Use Outside Label:
9. Finally, click on the Common Settings button, change the Chart Type to Donut and toggle the chart Legend:
10. Save the chart and the system will place it automatically on the Dashboard.
11. Now let's add a new chart. Click on the plus icon next to the Chart section in the right pane.
12. Click on the OrderDate dimension, Sales and Profit measures and select Line Chart as the type of chart to be created.
13. Change the aggregation type for the Sales and Profit measures to AVG:
14. Change the Granularity for OrderDate to Month:
15. Now let's add data labels to our line chart. Click on the Data Label settings option, enable Show Data Label, select Series Value and choose Above for Position:
16. Click on Save and drag the new chart below the first pie chart.
17. Now let's add another one. Click on the plus icon next to the Chart section title. Select Text Table as the chart type.
18. Drag the City column to the second dimension slot and the Sales column to the measure slot:
19. Change the order of Sales items to Descending:
20. Let's format our numbers now to look more friendly. Click on the Number Format button, change Format to Currency and choose USD as the currency. Select Automation for Number Abbreviations.
21. Let's also color the numbers based on their highest value. Click on the Color Setting button and choose Measure from the dropdown. Click on the color selector and choose the red scheme from the 9 steps category.
22. Click on the Common Setting icon and set Color Style to background:
24. Finally, Save the chart and drag it on the right side of the dashboard.
25. Click on the Done button on the top ribbon to save our dashboard.
While viewing a dashboard, you can drill-down into various items by clicking on them into each widget.
For example, to see what happened in New York City, click on the New York City row in the right table widget and the dashboard will refresh to show only information related to New York City.
26. Now let's create a new dashboard with the salesdata datasource and name it "Sales Geography". Add a new chart to the dashboard and choose Map View for the chart type. Add the Geolocation dimension and Sales measure to the chart:
27. Click on the Layer Settings button and change the Layer Type to Heatmap, Hexagon or Cluster to see the different type of spatial aggregations.
28. Thank you for your patience and keep playing around to discovery other BigConnect Discovery features like the Workbench, Notebooks, Data Catalogue, Collaboration features and more.
Last modified 1yr ago