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)

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

A screenshot of an area chart type example that shows average CPU usage.

The area chart type 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.

Use an area chart when you want to	View the measurement of a single attribute over a period of time that you specify.
Avoid an area chart when you want to	View a chart that contains more than a single attribute. The line chart type is a good option; you can view multiple attributes over a time range that you specify. The bar chart type gives you a comparison view between the attributes; it uses a total of each attribute for the time range you choose.

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

A screenshot of a bar chart example that shows a comparison of browser usage.

The bar chart compares a related set of values.

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.

Use a bar chart when you want to	Create a chart that reports the total count of the data for a category over the specified time range. With the bar chart format, you can easily see the differences between each category.
Avoid a bar chart when you want to	Include measurements over a period of time. This chart type is not available when using a TIMESERIES clause. The line chart type is a good option for this visualization.

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.

Billboards can also be used 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.

The billboard chart displays a single value in a large, easy to read font. The background color changes based on your thresholds.

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.

팁

Threshold fields will only be available for dashboards that present a single value. For instance, the use of a COMPARE WITH clause will disable the threshold settings in the edit pane.

Uses

Use a billboard chart when you want to	Monitor values, including thresholds for Critical and Warning states, in a large, easy-to-read font.
Avoid a billboard chart type when	You want to view data over a period of time. The line chart type is a better option for viewing data by time.

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

A screenshot of an example bullet chart that shows the total number of unique pageviews out of a total number of pageviews.

The bullet chart displays a value compared against a limit.

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.

팁

The bullet chart type can only be used for charts that present a single value. For example, a query using COMPARE WITH will not list bullet as an option.

Use a bullet chart when you want to	View a chart uses a large, easy-to-read font that helps you view the value quickly.
Avoid a bullet chart when you want to	View data that has more than a single value. Bar and line chart types may be better choices in this instance.

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)?”

A screenshot of an example NRQL funnel query that displays a count of the users who registered for an account, visited a feature-compare page, and upgraded.

Sample funnel chart for a shopping site.

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 is not 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

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.

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

A screenshot of an example heatmap chart showing a pageview breakdown by operating system.

The heat map is a faceted distribution of values where color density represents the values displayed.

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.

Use a heat map when you want to	View tabular data for multiple attributes over a period of time, with gradated colors instead of numerals.
Avoid a heat map when you want to	View the exact data for multiple attributes over a period of time. A line or bar chart may be an option.

A screenshot of an example histogram showing pageview duration over time.

A histogram charts frequency across a series of value ranges.

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

Use a histogram when you want to	Chart frequency of data over a series of values.
Avoid a histogram when you want to	View the same data as heat map or to view the data without the time buckets.

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
  }
}

...

A line chart is useful for spotting trends over time.

A screenshot of an example line chart showing pageview count over time.

The line chart plots a timeseries for an attribute.

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.

Use a line chart when you want to	Create a chart that lets you easily view trends over a specified time frame.
Avoid a line chart when you want to	View your data using one or more attributes and see the values associated with each attribute. A bar or table chart is an option.

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.

The pie chart maps a set of values as a portion of a whole.

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%.

Use a pie chart when you want to	Create a chart that that graphically presents a set of values as they relate to a whole. A pie chart is effective for displaying values for a time period, for example.
Avoid a pie chart when you want to	Present data that does not use values as they related to a whole, such as with line charts.

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.

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). 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.

Other features of tables:

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 export a table chart to a CSV file. Details on this functionality:

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.
The table chart type, with its tabular format, works well for dashboards, as well as sharing the data for use in reports, for example.

Use a table when you want to	Create a text-compatible table consisting of your data. This table can be shared and used in reports or imported into spreadsheets, for example.
Avoid a table when you want to	View your data in a graphical format.

Area chart

Bar chart

Billboard

팁

Bullet chart

팁

Funnel

Heat map

Histogram

JSON

Line

Pie chart

Pie chart percentage problem

Table