Chart types

When building a query using the query builder, there are a variety of types of visualizations for presenting your data in the optimal format.

This table contains all chart types. Whether a chart type is available for your query will be dependent on:

• The type of query (some query results don't match some chart types)
• The query mode (basic or advanced)

Area chart

An area chart shows a time series for a single attribute.

To enable an area chart, add a FACET to your query.

You can use facet to add up to five attributes, separated by commas. Use the TIMESERIES function when you use a FACET in an NRQL query.

Bar chart

The bar chart shows comparisons among the categories returned in the query results.

To enable a bar chart, add a FACET clause to the query.

You can use FACET with up to 5 different attributes, separated by commas.

Billboard

The billboard chart type presents results as a single, highly-visible value. It's an option when your query returns a single number, such as from this query:

SELECT uniqueCount(session) FROM PageView since 1 day ago

A billboard chart is useful for single-stat queries like the Apdex function. Use billboards to monitor values at a glance.

You can also use billboards to compare a single value across states. For instance, use the COMPARE WITH clause to compare the same attribute across two different time frames.

To enable the data for a billboard chart, search for a single value. You can view multiple values by adding a facet from the dropdown.

Thresholds

Optionally, you can set Warning and/or Critical thresholds on billboards. Once set, if your service is working correctly, the chart's full box is green. When a threshold is crossed, the chart turns yellow (warning) or red (critical). Access the threshold fields on the right side of the chart edit pane in the query builder.

Here are some tips to help you with thresholds:

• For percentages, use a decimalized format. For example, if your threshold is 99.5%, divide by 100 and set the value to 0.995.
• If you want to highlight a chart when a value falls below a given value rather than above it, set the Warning threshold higher than the Critical threshold.

Uses

Bullet chart

Use the bullet chart type to represent progress to a target or goal, called the limit. To use this chart type, run a query that returns a single number, such as:

SELECT uniqueCount(session) FROM PageView SINCE 1 day ago

Bullet defaults to a limit that is twice the resulting query value. To adjust the limit, use the Limit field to enter another number. The bullet will adjust accordingly.

Funnel

Use funnels to evaluate a series of related data types that make up the steps toward a specified goal. A funnel allows you to ask, “Of the people that did (A), how many also did (B), and of the people that did (A) and (B), how many also completed (C)?”

In this example, you want to determine that of all the visitors that viewed the landing page, how many of them:

• Browsed the products.

• Then added one or more products to a shopping cart.

• And finally, purchased the item(s) in the shopping cart.

  Unlike other queries, you must use a SELECT funnel() statement in your query; the funnel chart type isn't selectable from any other type of query.

  Additionally, the funnel chart type is only available for queries with eight steps or fewer. A funnel query that includes more than eight steps will only return the results as JSON.

  SELECT funnel(session,
      WHERE pageUrl like '%/index.html' AS 'Landing',
      WHERE pageUrl LIKE '%/browse/phones' AS 'Browse ',
      WHERE pageUrl LIKE '%/shoppingcart' AS 'Cart',
      WHERE pageUrl LIKE '%/checkout' AS 'Purchased')
      FROM PageView where appName = 'WebPortal' AND duration < 1
      SINCE 1 week ago

  コピー

  Use a funnel when you want to	You want to visually evaluate the paths that visitors take after they hit a page that you designate as a landing page.
  Avoid a funnel when you want to	You only want counts for the attributes in the WHERE clause. You will not be able to choose funnel unless you use SELECT funnel in your query.

Heat map

The heat map displays data in a tabular format using color gradations to show frequency in specific time buckets.

To enable a heat map, use the Histogram() function with a numeric attribute and use FACET to add up to five attributes, separated by commas.

See Histograms: View data distribution for more detailed information about histogram-based charts.

Histogram

Use histograms to show the distribution of data across time buckets. See Histograms: View data distribution for more detailed information about histogram-based charts.

JSON

Use the returned JSON to review the raw data returned and to help develop integrations with New Relic. To implement integrations, use remote queries to return raw JSON as a service.

To use this, select the Share menu and click Copy JSON.

Example NRQL query:

SELECT uniqueCount(name) FROM Transaction
WHERE http.statusCode = '200'
TIMESERIES
FACET appName
LIMIT 10
SINCE 24 hours ago

Example JSON response:

{
  "metadata": {
    "contents": {
      "messages": [],
      "timeSeries": {
        "messages": [],
        "contents": [
          {
            "function": "uniquecount",
            "attribute": "name",
            "simple": true
          }
        ]
      },
      "bucketSizeMillis": 1800000
    },
    "eventTypes": [
      "Transaction"
    ],
    "eventType": "Transaction",
    "openEnded": true,
    "messages": [],
    "beginTimeMillis": 1556226069396,
    "endTimeMillis": 1556312469396,
    "beginTime": "2019-04-25T21:01:09Z",
    "endTime": "2019-04-26T21:01:09Z",
    "guid": "5238a2c6-328d-a5ff-9185-7f3f04c42d1e",
    "routerGuid": "1aea2ec5-0be8-3a60-7554-851e1b016eb0",
    "rawSince": "24 HOURS AGO",
    "rawUntil": "NOW",
    "rawCompareWith": "",
    "facet": "appName",
    "offset": 0,
    "limit": 10
  },
  "facets": [
    {
      "name": "RPM Combined Production",
      "beginTimeSeconds": 0,
      "endTimeSeconds": 0,
      "timeSeries": [
        {
          "results": [
            {
              "uniqueCount": 415
            }
          ],
          "beginTimeSeconds": 1556226069,
          "endTimeSeconds": 1556227869,
          "inspectedCount": 4589471
        },
        [
          {
            "uniqueCount": 1714
          }
        ],
        "beginTimeSeconds": 1556310669,
        "endTimeSeconds": 1556312469,
        "inspectedCount": 53341469
      }
    ],
    "total": {
      "results": [
        {
          "uniqueCount": 2651
        }
      ],
      "beginTimeSeconds": 0,
      "endTimeSeconds": 0,
      "inspectedCount": 5053841114
    }
  },
  "performanceStats": {
    "inspectedCount": 6088068932,
    "responseTime": 8077
  }
}

...

Line

A line chart is useful for spotting trends over time.

When working in basic query mode, select a single attribute you can plot over time.

When using NRQL, SELECT one or more attributes you can plot over time and then use the TIMESERIES function.

Pie chart

A pie chart is the graphic representation of the table. It appears as an option for some functions when the FACET clause is used. To use this chart type, run a query that returns one value for each of a set of attributes.

To enable pie chart, add a FACET from the dropdown.

You can add only one FACET in basic query mode. If you're using NRQL, you can use FACET to add up to 5 attributes, separated by commas, and also include the TIMESERIES function.

Pie charts are unavailable for the following aggregator functions: average, apdex, min, max, percentage, and percentile. For uniqueCount, percentages may add up to more than 100%.

Pie chart percentage problem

When using a pie chart with the uniqueCount aggregator function, the percentages can add up to more than 100%. This is because the attributes being uniquely counted may be present in multiple facets. For example, in the query SELECT uniqueCount(user) FROM PageView FACET appName, a single unique user may use multiple apps. Each of these users are included in the unique value for each of the appropriate facets (apps), but the total number of unique users won't change.

To solve this, use a bar chart or table to provide a more accurate visualization of uniqueCount data.

Table

A table is the standard view for a set of values related to a specific attribute (for example, the count of all page views by user agent name, log details, etc.). By default, all tables are sortable by any one column.

If you are using the FACET clause, the results will normally be returned as a table.

You can use a table as a shortcut to other dashboards. If a table is faceted by attribute, selecting an entry takes you to an existing dashboard filtered by that attribute.

You can also export a table chart to a CSV file.

• If a filter is applied to a dashboard, the download will only include the subset of data based on the filter.
• If a specific fixed time range is selected, the CSV includes the data for that time period.
• If you query with multiple attributes, the export includes columns for all FACETs in the query.
• For charts with a time range dependent on the present moment (for example, since 30 minutes ago), the data in the CSV is generated from the present moment, not from when the chart was initially generated.