This blog post describes the basics of time series analysis in Azure Data Explorer.
Azure Data Explorer (ADX) is a service designed for fast data exploration. It provides instant insights into large datasets to analyze performance, identify trends and anomalies, and troubleshoot problems.
The collection of telemetry data is performed from cloud services or IoT devices. The analysis can be performed on sets of time series for selected metrics to find a deviation in the pattern of the metrics relative to their typical baseline patterns.
ADX provides native support for the creation, editing, and analysis of time series in near real-time.
Types of the time series analysis
First, the original telemetry table is partitioned and transformed to a set of time series using the make-series operator. ADX provides then the following capabilities for the analysis:
- Filtering – noise reduction, smoothing, change detection, and pattern matching
- Regression analysis – trend change detection in streamed data
- Seasonality detection – automatic detection or validation of the seasonal or periodic patterns in each time series
- Element-wise functions – perform arithmetic and logical operations between several time series
The complete set of functions for the time series analysis is available in the Microsoft documentation section.
Sample time series analysis query
The following time series analysis discovers periodic patterns and decreasing trends with series_periods_detect and series_fit_line functions:
In the above query 18,339 time series of web service traffic are analyzed and extracted with a periodic pattern.
- get started with Azure Data Explorer in preview
- step-by-step walkthrough of time series analysis capabilities
- Azure Data Explorer Technical 101 blog