15.02.2023 Area A Data Modeling

[hampusnasstrom]

Agenda

Following last weeks discussion on the specification of array data we discuss:

• Implementation of array data

Meeting notes

For plotting/previewing large arrays we could request a reduced set from the API (Theodore working on something similar, options dict for each property)
Ideally, give back data <X MB.

Order of retrieval

1. Load statistical parameters
2. Load reduced preview
3. Load full array

Two different goals

1. View it in ELN (ElnComponent)
2. Search-ability (Nested elastic annotations)

Seperate feature:
Generate table view like the plot gitlab issues #1269 #933

Implementation

Both NumericalArray and ContextArray should be implemented in nomad/metainfo/metainfo.py.

Implemented as either 'new Quantity class', or 'subclass of Quantity' or 'type'. Which will be discussed in issue/MR.

Data reference in order of implementation

1. To a data path
2. To H5 (Alvin)
3. (later) reference to csv (possibly through h5)

Tasks

• Hampus creates issue
• Hampus makes draft of numerical and context array

Next week

• Reference "roles" (Workflow tasks, Ensemble components)

[hampusnasstrom]

Issue on gitlab: https://gitlab.mpcdf.mpg.de/nomad-lab/nomad-FAIR/-/issues/1352