4. Descriptive Statistics

Use Cases

Descriptive Statistics is a tool that provides a detailed look at the behavior of a particular parameter overall and over time. This tool is most helpful later in the exploratory process, after you have identified a parameter to analyze in detail. This is a good tool to follow Variance Analysis or the Event Timeline.

Specifications

Descriptive Statistics is a univariate analysis that you can perform on continuous and categorical fields on the Cycle model. You can select multiple machines of a given type. This analysis does not support the Part model because, when analyzing a single field, the extra context provided by the Part model is not required, and the Cycle model has higher granularity.

Analysis

Descriptive Statistics is an exploratory tool that computes metrics about a particular parameter to analyze its behavior overall and over time. For continuous fields, the tool plots a histogram of the parameter and computes a table of summary statistics that includes Count, Standard Deviation, Minimum, and Maximum. For categorical fields, the tool computes the count and distribution of each category value. The tool also plots the parameter’s distribution by week, which can highlight drift over time, and the number of records from each day, which can highlight missing data.