Page Comparison

...

1. Create a user

   1. Note: you can only create one user per email address
   1. Note: once you create a user, only that user can edit their password. Administrators cannot reset passwords in Superset's UI currently, it needs to be done through the command line
      1. https://github.com/apache/incubator-superset/issues/4518
1. Security → List Users
2. Click the plus sign (+) in the top right
3. Fill out the required fields
4. Save

2. Create a user role

1. Security → List Roles
2. Specify the security points

3. Access your datasource

   1. Sources → Druid Datasources
   1. Sources → Tables
1. Druid
2. PostgreSQL (or other RDBMS)

4. View and edit columns - incl. column labels

   1. Click the Edit Record button to the left of the column’s name
   2. Druid columns
      1. {
      2.  "type" : "default",
      3.  "dimension" : "actual_dimension_name",
      4.  "outputName": "Desired Name",
      5.  "outputType": "STRING"
      6. }
   1. Lst this in the Dimension Spec Json field:
   3. RDBMS columns
   1. Fill in the “Verbose Name” field
1. Within each datasource (see #3 above), click List Columns
2. Specify whether you want the column to be available for Grouping By within queries or Filtering within queries/dashboards
3. To edit a column’s name

5. Create basic metrics

1. Within each datasource (see #3 above), click List Columns
2. For any relevant column, click the checkboxes for Count Distinct, Sum, Min, or Max. This will auto-generate metrics on the List Druid Metrics tab. Currently, Sum, Min, and Max are selected by default for any numeric-type column

6. Create complex metrics

   1. RDBMS - option 1
      1. Name (Metric field) (required)
      2. Description (optional)
      3. Verbose Name (optional)
      4. Type (required)
      5. SQL expression (required)
      1. This is a simple SQL expression between two or more columns, it supports basic aggregation functions like AVG, subtraction, division. It does not require SELECT or FROM clauses, and does not support more advanced clauses like WHERE, OVER, or setting variables.
   1. Click the plus sign (+) to the top left
   2. Fill in the following important fields when appropriate:
   3. Click Save
   2. RDBMS - option 2
      1. Specify your time column
      2. Specify any columns that should be treated as dimensions
      3. Specify any aggregate functions that should be applied
      4. Select a Chart Type (this can be changed once you’re within the Chart Editor)
      5. Specify a Datasource Name
      1. Note this means any Charts built off of this new datasource will not be affected by filters applied to the “parent” datasource (where you’re selecting from in the SQL query)
      6. See the Create/edit Charts section (#7) below
   1. SQL Lab → SQL Editor
   2. Write a full SQL expression, including SELECT, FROM, WHERE, etc. clauses.
   3. Run Query shows your output in tabular format
   4. When your query is complete, click the Visualize button and:
1. Within each datasource (see #3 above), click List Metrics

7. Create/edit Charts

   1. Edit Visualization Type
   2. Edit time granularity and time range (Since and Until)
   3. Specify which metrics to display
   4. Specify which dimensions/columns to Group By
   5. Edit Number format - this supports free text entries, so typing “100%” will convert the results into percentages
   6. Apply any default filters (based on dimensions) or result filters (based on metric values)
1. Click the Charts button at the top
2. Click the Edit record button to the left of the chart name to view a description, owners, associated dashboards, and parameters JSON (including default filters) for the Chart
3. Click the Chart name to open the Chart Editor. Note that each visualization type has its own parameters, but we will cover the essentials here:

8. Create/edit a dashboard

   1. Click the plus sign (+) in the top right
   2. Specify a title and owners
   3. Your dashboard is now available in the Dashboards menu, follow steps 8c below
   1. Click the Edit record button to the left of the dashboard name to view associated charts, owners, CSS and JSON metadata for the dashboard.
   2. Click on the dashboard’s name to open it
   3. Click the Edit Dashboard button in the top right to adjust positioning of charts, to add/remove charts, or to set a refresh schedule (found under the Actions dropdown)
   4. Click Save
   5. Click Switch to View Mode
1. Click Dashboards at the top
2. To create a dashboard
3. To edit a dashboard

9. Drill into a Chart

1. Dashboards → dashboard name
2. In the top right of a Chart, click the Explore Chart button (right-pointing arrow). This opens the Chart Editor (see #7c above)
10. Apply a default filter to a dashboard
    1. Dashboard list → edit dashboard
    2. In the JSON Metadata field, enter this: "default_filters": "{\"id\": {\"dimension\": [\"value\"]}}"
       1. id is the NUMERIC id of the filter slice you want to filter by
       2. dimension is the dimension name (i.e. column name) that you want to filter by
       3. value is the dimension value that you want to filter to
       4. "default_filters": "{\"1705\": {\"country\": [\"Kenya\"]}}" this applies a default filter to a dashboard using chart id 1705, and it will only return results where the country column is equal to Kenya
11. Apply default colors to charts
    1. Dashboard list → edit dashboard
    2. In the JSON Metadaata field, enter this: "label_colors": {"Value1": "color1", "value2": "color2"...}
       1. Color must be expressed using hex codes