ThingWorx Analytics Training: Module 3 Part 2
This video continues Module 3: Data Profiling of the ThingWorx Analytics Training videos. It describes metadata, and how it is used to ensure that your data is handled appropriately when running Signals, Profiles, Training, Scoring, and other jobs inside ThingWorx Analytics.
This video does not contain any hands-on exercises, and you don't need access to a running ThingWorx Analytics environment.
This video has chapters — use the chapter menu in the player to jump straight to any section below.
Chapter Summaries
0:00 – Data Profiling & Metadata Basics: JSON Format for ThingWorx Analytics
Explains how data profiling helps create or validate metadata—data that describes a ThingWorx Analytics data set—which must accompany every loaded data set and is defined using a JSON file format.
0:30 – Field Name & Values Parameters: Ordinal and Categorical Requirements
Covers the required field name that must exactly match the source feature, plus the values parameter listing possible entries for ordinal (ordered) and categorical opt types, omitted for Boolean or continuous fields.
1:00 – Range Parameter: Min/Max Bounds for Continuous Fields
Describes the optional range parameter for continuous opt types, holding informational min and max values that don't restrict data but can drive a mashup display of the field's possible value range.
1:21 – Data Type Parameter: String, Boolean & Integer Options
Details the required data type parameter—string, Boolean, or integer—and warns that overriding the type ThingWorx Analytics auto-selects triggers an error when incoming data doesn't match the selected type.
1:46 – Opt Type Overview & Categorical/Boolean Types: Finite Values, True/False Flags
Introduces the required opt type parameter that defines how data behaves, then explains categorical types (finite grouped values like gas/diesel fuel, with no inherent order) and Boolean true/false flags such as a serviced-in-30-days indicator.
2:54 – Ordinal & Continuous Opt Types: Ordered Values and Numeric Fields
Explains ordinal opt types with a finite, inherently ordered value set (low, medium, high) and continuous opt types used for numeric fields with double, integer, or long data types.
3:22 – Information & Text Opt Types: Serial Numbers & Free-Form Text Analysis
Covers the information opt type for display-only data like serial numbers excluded from analytic services, and the text opt type for free-form comments or notes processed with text analysis in descriptive and predictive jobs.
4:02 – Temporal & Entity ID Opt Types: Time Series Data Requirements
Describes time series data requiring exactly one temporal field (ordering the data rows) and one entity ID field (identifying an asset such as a machine serial number), previewed ahead of the later time series module.
4:45 – Opt Type Performance Impact: Categorical vs Continuous Errors
Warns that changing an auto-detected opt type incorrectly—such as defining a high-cardinality numeric field as categorical instead of continuous—forces the system to treat each value as its own category, drastically increasing processing time.
5:17 – Time Sampling Interval & Is Static Flag: Time Series Pivoting
Explains the required time sampling interval between adjacent temporal rows and the is static flag, which stops fields like a device serial number or ZIP code from undergoing time series pivoting to improve performance.
6:04 – Automated Metadata Creation in ThingWorx Analytics 8.4
Notes that authoring JSON metadata manually is complex, and that ThingWorx Analytics 8.4 automated metadata creation during data set setup—letting you bypass, review, modify, or download the inferred JSON metadata.
6:46 – Detect Metadata Service: Analytics Builder & Data Microserver
Shows metadata auto-generation in both Analytics Builder and the ThingWorx services, where the detect metadata service in the data microserver infers metadata and builds an analytics data set metadata info table for review.
7:14 – Uploading Metadata in Analytics Builder: Upload Checkbox & File Selection
Walks through the Analytics Builder data upload screen—using the Upload Metadata checkbox and Choose File to supply a JSON file, Review Uploaded Metadata to edit it, or leaving the box unchecked so ThingWorx infers the metadata.
7:49 – Reviewing & Editing Metadata: Opt Type Dropdowns & Values Editing
Demonstrates the review/modify screen shown for uploaded or inferred metadata, using the opt type and data type dropdowns and editing the values parameter—one value per line, with ordinal values entered in order.