Creating and Managing Views#

Views enable you to aggregate data from various datasets, transform the data, and share the resulting data with applications. A view is a virtual dataset created by joining multiple datasets. The data aggregation and transformation enable you to enrich data for applications.

Here we’ll create a view of cars and related car accidents by joining a datasets containing each type of data.

Note

You can apply these principles to combine data from any datasets that can be joined on equivalent indexed properties.

Create example data files#

  1. Create a plain text data file called cars on your local machine from the following example car data (it’s in CSV format).

    Note

    The product supports CSV files that use the following common data delimiters: comma (,), tab, or pipe (|) characters. JSON and JSONL files are also supported.

    Data

    vin,make,model,year,current_date,purchase_date,estimated_value,mileage,owner_count
    XV859643N98D98E7C,Chevrolet,Camaro,2020,2020-03-27,2020-03-13,45955.00,32000,2
    SD089VN7678997566,Ford,F-150,2022,2020-03-27,2022-01-11,38650.00,8900,1
    59ADFG60929087DAH,Toyota,Prius,2018,2020-03-27,2019-09-23,22876.00,76000,1
    

    Example Command

    These commands create the data file on Linux, MacOS, and Windows.

    echo "vin,make,model,year,current_date,purchase_date,estimated_value,mileage,owner_count
    XV859643N98D98E7C,Chevrolet,Camaro,2020,2020-03-27,2020-03-13,45955.00,32000,2
    SD089VN7678997566,Ford,F-150,2022,2020-03-27,2022-01-11,38650.00,8900,1
    59ADFG60929087DAH,Toyota,Prius,2018,2020-03-27,2019-09-23,22876.00,76000,1" \
    >>CARS
    
    (
    echo vin,make,model,year,current_date,purchase_date,estimated_value,mileage,owner_count
    echo XV859643N98D98E7C,Chevrolet,Camaro,2020-03-27,2020,2020-03-13,45955.00,32000,2
    echo SD089VN7678997566,Ford,F-150,2022,2020-03-27,2022-01-11,38650.00,8900,1
    echo 59ADFG60929087DAH,Toyota,Prius,2018,2020-03-27,2019-09-23,22876.00,76000,1
    )>CARS
    
  2. Create a file called car_accidents from the following example car accident data.

    Data

    city,date,est_damage,state,vin
    Hoboken,2021-01-01,18000,NJ,W9SD8B762APZKE003
    Harrisburg,2022-02-03,1500,PA,SD089VN7678997566
    

    Example Command

    These commands create the data file on Linux, MacOS, and Windows.

    Linux/MacOS

    echo "city,date,est_damage,state,vin
    Hoboken,2021-01-01,18000,NJ,W9SD8B762APZKE003
    Harrisburg,2022-02-03,1500,PA,SD089VN7678997566" \
    >>car_accidents
    

    Windows

    (
    echo city,date,est_damage,state,vin
    echo Hoboken,2021-01-01,18000,NJ,W9SD8B762APZKE003
    echo Harrisburg,2022-02-03,1500,PA,SD089VN7678997566
    )>car_accidents
    

Create datasets from the files#

  1. Click Create a Dataset at the top of the console or from the Datasets page. The Create a Dataset page appears.

  2. In the Use a new data source section, choose File in the source type, drag your cars file onto the page, drop it, and click Parse data. The file uploads and the Edit schema interface appears.

  3. In the Edit schema interface, modify the schema to specify the dataset ID, property characteristics, and Unique Index using the following values.

    Important

    The _system prefix (case-insensitive) is reserved for Apperate system tables and columns. You are forbidden to prefix dataset IDs, view IDs, and dataset property names with _system (case-insensitive).

    Dataset ID: CARS

    Unique ID related Properties

    Required

    Allow null

    Property

    Index

    x

    vin (string)

    Primary

    x

    make (string)

    Secondary

    x

    current_date (date)

    Date

    Remaining Properties

    Required

    Allow null

    Property

    x

    model (string)

    x

    year (integer)

    estimated_value (number)

    x

    mileage (integer)

    owner_count (integer)

    x

    purchase_date (date)

    When you’re done editing, click Create dataset. The CARS dataset overview appears.

    You’ve created the CARS dataset. It’s time to create one for car accidents.

  4. Start creating a CAR_ACCIDENTS dataset by clicking Create a dataset.

  5. Drag your car_accidents file onto the page, drop it, and click Parse data. The sample data from the file uploads and the Edit schema interface appears.

  6. In the Edit schema interface, modify the schema to specify the dataset ID, property characteristics, and Unique Index using the following values.

    Dataset ID: CAR_ACCIDENTS

    Unique ID related Properties

    Required

    Allow null

    Property

    Index

    x

    vin (string)

    Primary

    x

    state (string)

    Secondary

    x

    date (date)

    Date

    Remaining Properties

    Required

    Allow null

    Property

    x

    city (string)

    est_damage (number)

    When you’re done editing, click Create dataset. The CAR_ACCIDENTS dataset overview appears.

    We’re ready to create a view by joining data form the CARS and CAR_ACCIDENTS datasets.

Example: Creating a view#

Follow these steps to create the example car accidents view:

  1. Go to your CARS dataset’s Database page (next to the dataset Overview tab).

  2. In the Database page’s SQL editor, run a query that joins the datasets.

    SELECT a.vin, a.make, a.model, a.mileage, c.date, c.city, c.state, c.est_damage
      FROM MY.`CARS` a 
      JOIN MY.`CAR_ACCIDENTS` c 
      ON c.`vin` = a.`vin`
    

    Data from both datasets appears in the Results table below the editor.

    Important

    WHERE clauses and ON clauses must only operate on indexed properties (columns). See the Unique Index components here.

  3. Create a view from the data results by clicking the Create view button at the top right of the results table. The Create a view using SQL dialog appears.

  4. Name your view (or go with the unique auto-generated name) and click Create View. Your new view’s Overview page appears.

    Important

    The _system prefix (case-insensitive) is reserved for Apperate system tables and columns. You are forbidden to prefix dataset IDs, view IDs, and dataset property names with _system (case-insensitive).

    View pages are equivalent to a dataset pages, and views also appear in the datasets navigation tree.

  5. Execute the view’s example request in your browser clicking the Example Request URL. A browser tab opens and results like this appear:

    [{"city":"Harrisburg","date":"2022-02-03","est_damage":1500,"make":"Ford","mileage":8900,"model":"F-150","state":"PA","vin":"SD089VN7678997566"}]
    

Congratulations on combining data from two datasets into a view for your app to use.

Note

The Example Request returns just one record by using the LIMIT 1 condition. In your app, make sure to lift that limit by removing that condition from your URL’s SQL query.