visions and ydata-profiling
ID visions
detector
visions is a Python library for semantic type detection and inference on tabular data. It organizes types, such as URL, Path, File, Image, UUID, IPAddress, email address, Geometry, Date and DateTime, in typesets and tests which type fits a series. ydata-profiling (formerly pandas-profiling) uses it to infer column types for its data profile reports. The types are Python-level kinds of values, many are broader than registry datatypes. The supported types cover only registry datatypes that match a documented type.
- Website
- https://github.com/dylan-profiler/visions
- Supported types
- 9
Supported semantic types
| ID | Name | Categories | Countries | Description |
|---|---|---|---|---|
| date | Date | datetime | global | Calendar date. The regexp covers the ISO 8601 extended form YYYY-MM-DD; other notations, such as DD.MM.YYYY, are patterns of this datatype. |
| datetime | Date and time | datetime | global | Date and time of day. The regexp covers the ISO 8601 extended form YYYY-MM-DDThh:mm[:ss[.fff]] with an optional time zone (Z or an... |
| duration | Time duration | datetime | global | The amount of time elapsed between two events |
| Email address PII | common, internet, pii | global | Address of an electronic mailbox in the form local-part@domain, defined in section 3.4.1 of RFC 5322. The local part names the mailbox,... | |
| filepath | File path draft | common | global | Location of a file in a file system, written as directory names joined by a separator and ending with the file name. Unix-like systems... |
| ipaddr | IP address (IPv4 or IPv6) PII | internet, pii | global | Address of a network interface in the Internet Protocol: a 32-bit IPv4 address in dotted-decimal form (192.0.2.1) or a 128-bit IPv6... |
| time | Time draft | datetime | global | Time of the day, like 12:00 or 10PM or 11:05 or 18:01:27 |
| url | Uniform Resource Locator (URL) | common, internet | global | Address of a resource on a network that also says how to retrieve it: a scheme such as https, a host, an optional port, a path, a query... |
| uuid | Universally unique identifier (UUID) | identifiers | global | 128-bit identifier that systems generate without a central registry, defined by RFC 9562 (2024), which replaced RFC 4122. Microsoft... |
Use this record
- JSON: API v1 · legacy shape
- Improve it: edit
data/tools/detectors/visions.yamlon GitHub