Skip to main content

Python dataclasses for the OME data model

Project description


License Version CondaVersion Python Version Tests Docs

A pure-python implementation of the OME data model

ome_types provides a set of python dataclasses and utility functions for parsing the OME-XML format into fully-typed python objects for interactive or programmatic access in python. It can also take these python objects and output them into valid OME-XML. ome_types is a pure python library and does not require a Java virtual machine.

Note: The generated python code can be seen in the built branch. (Read the code generation section for details).

📖   documentation


from pip

pip install ome-types

from source

git clone
cd ome-types
pip install -e .


convert an XML string or filepath into an instance of ome_types.model.OME

(The XML string/file will be validated against the ome.xsd schema)

from ome_types import from_xml

ome = from_xml('tests/data/hcs.ome.xml')

extract OME metadata from an OME-TIFF

from ome_types import from_tiff

ome2 = from_tiff('tests/data/ome.tiff')

manipulate the metadata via python objects

Both from_xml and from_tiff return an instance of ome_types.model.OME. All classes in ome_types.model follow the naming conventions of the OME data model, but use snake_case attribute names instead of CamelCase, to be consistent with the python ecosystem.

In [2]: ome = from_xml('tests/data/hcs.ome.xml')

In [3]: ome
    images=[<1 Images>],
    plates=[<1 Plates>],

In [4]: ome.plates[0]
    name='Control Plate',
    wells=[<1 Wells>],

In [5]: ome.images[0]
    name='Series 1',
        bin_data=[<1 Bin_Data>],
        channels=[<3 Channels>],
    description='An example OME compliant file, based on Olympus.oib',

Objects can be removed, or changed

In [6]: from ome_types.model.simple_types import UnitsLength

In [7]: from import AcquisitionMode

In [8]: ome.images[0].description = "This is the new description."

In [9]: ome.images[0].pixels.physical_size_x = 350.0

In [10]: ome.images[0].pixels.physical_size_x_unit = UnitsLength.NANOMETER

In [11]: for c in ome.images[0].pixels.channels:
             c.acquisition_mode = AcquisitionMode.SPINNING_DISK_CONFOCAL

Elements can be added by constructing new OME model objects

In [12]: from ome_types.model import Instrument, Microscope, Objective, InstrumentRef

In [13]: microscope_mk4 = Microscope(
             manufacturer='OME Instruments',
             model='Lab Mk4',

In [14]: objective_40x = Objective(
             manufacturer='OME Objectives',

In [15]: instrument = Instrument(

In [16]: ome.instruments.append(instrument)

In [17]: ome.images[0].instrument_ref = InstrumentRef(

In [18]: ome.instruments
       manufacturer='OME Instruments',
       model='Lab Mk4',
    objectives=[<1 Objectives>],

export to an OME-XML string

Finally, you can generate the OME-XML representation of the OME model object, for writing to a standalone .ome.xml file or inserting into the header of an OME-TIFF file:

In [19]: from ome_types import to_xml

In [20]: print(to_xml(ome))
<OME ...>
    <Plate ColumnNamingConvention="letter" Columns="12" ID="Plate:1" ...>
    <Instrument ID="Instrument:1">
        <Microscope Manufacturer="OME Instruments" Model="Lab Mk4" SerialNumber="L4-5678" />
        <Objective Manufacturer="OME Objectives" Model="40xAir" ID="Objective:1"
        NominalMagnification="40.0" />
    <Image ID="Image:0" Name="Series 1">
        <Description>This is the new description.</Description>
        <InstrumentRef ID="Instrument:1" />
        <Pixels ... PhysicalSizeX="350.0" PhysicalSizeXUnit="nm" ...>
            <Channel AcquisitionMode="SpinningDiskConfocal" ...>

Code generation

The bulk of this library (namely, the ome_types.model module) is generated at install time, and is therefore not checked into source (or visible in the main branch of this repo).

You can see the code generated by the main branch in the built branch

The script at src/ converts the ome.xsd schema into valid python code. To run the code generation script in a development environment, clone this repository and run:

python src/

As an example, the OME/Image model will be rendered as the following dataclass in ome_types/model/

from datetime import datetime
from typing import List, Optional

from pydantic import Field

from ome_types._base_type import OMEType

from .annotation_ref import AnnotationRef
from .experiment_ref import ExperimentRef
from .experimenter_group_ref import ExperimenterGroupRef
from .experimenter_ref import ExperimenterRef
from .imaging_environment import ImagingEnvironment
from .instrument_ref import InstrumentRef
from .microbeam_manipulation_ref import MicrobeamManipulationRef
from .objective_settings import ObjectiveSettings
from .pixels import Pixels
from .roi_ref import ROIRef
from .simple_types import ImageID
from .stage_label import StageLabel

class Image(OMEType):
    id: ImageID
    pixels: Pixels
    acquisition_date: Optional[datetime] = None
    annotation_ref: List[AnnotationRef] = Field(default_factory=list)
    description: Optional[str] = None
    experiment_ref: Optional[ExperimentRef] = None
    experimenter_group_ref: Optional[ExperimenterGroupRef] = None
    experimenter_ref: Optional[ExperimenterRef] = None
    imaging_environment: Optional[ImagingEnvironment] = None
    instrument_ref: Optional[InstrumentRef] = None
    microbeam_manipulation_ref: List[MicrobeamManipulationRef] = Field(
    name: Optional[str] = None
    objective_settings: Optional[ObjectiveSettings] = None
    roi_ref: List[ROIRef] = Field(default_factory=list)
    stage_label: Optional[StageLabel] = None

The documentation and types for the full model can be in the API Reference


To clone and install this repository locally (this will also build the model at src/ome_types/model)

git clone
cd ome-types
pip install -e .[autogen]

We use pre-commit to run various code-quality checks (isort, black, mypy, flake8) during continuous integration. If you'd like to make sure that your code will pass these checks before you commit your code, you should install pre-commit after cloning this repository:

pip install pre-commit
pre-commit install

If you modify src/, you may need to regenerate the model with:

python src/

Running tests

To run tests quickly, just install and run pytest. Note, however, that this requires that the ome_types.model module has already been built with python src/

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ome_types-0.4.0a3.tar.gz (64.4 kB view hashes)

Uploaded Source

Built Distribution

ome_types-0.4.0a3-py3-none-any.whl (166.2 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page