Welcome to FMU-dataio’s documentation

fmu-dataio is a specialized Python library for managing data in Fast Model Update (FMU) workflows. It focuses on exporting data while adhering to the FMU standards, which include both file and folder conventions, and rich metadata integration for various data consumers. The library is designed for consistent usage across all stages of FMU workflows, including ERT FORWARD_MODEL and pre-/post-processing jobs. fmu-dataio can be used both inside and outside RMS.

The purpose of fmu-dataio is to simplify data export and to add context to data produced by FMU workflows so that they can be used and understood also outside FMU. This is fundamental for enabling usage of the vast amounts of data produced by FMU without requiring significant manual intervention and repetitive work. The amount of context required is not possible to fit in a filename alone. Hence, fmu-dataio produces and attaches rich metadata to exported files.

In addition to the data export functions, fmu-dataio also contains the data model for FMU results.

While fmu-dataio represents a fair amount of simplification on its own, it is also a necessary investment for tapping into more simplification possibilities, such as management of results in Sumo, automated data pipelines to the Reservoir Experience Platform, centralized post-processing services, new and improved cloud-only version of Webviz and much more.