Is your feature request related to a problem? Please describe.
APM agents collect errors from an application and surface them in a tabular form. When a new type of error is captured it just adds that to the existing error table. Surfacing these new errors to make it more visible to the user that its new is immensely helpful in spotting them. These errors along with deployment annotations can help identify if a new deploy is potentially causing these issues.

In addition to the errors, the UI shows a latency distribution. It would be helpful to have an error rate chart that shows the aggregate error rate for that service and for a user to compare it to a week ago.

Describe the solution you'd like

• Add an additional error rate chart and allow a user to compare the error rate to a week ago.
• Annotate the error rate chart with versions.
• Surface new errors to the top of table and tag them so they stand out.
• Surface root cause of the issue by comparing the error with the other Transactions of the same type.
• Annotate the error chart to show triggered alerts.
• Add compare to a week ago to the error rate chart.

Additionally, we could also have an option on the error table to compare errors to a week ago and show only the errors that started occurring in the last week.

Design issue: #264
Error rate chart implementation issue: elastic/kibana#65619