omero-parade 0.1.3

OMERO.parade

An OMERO.web app for filtering Data in OMERO.web centre panel.

For full details see SUPPORT.md.

Requirements

• OMERO 5.4.0 or newer.

Note that OMERO.parade is compatible with OMERO 5.4.0 but it works best with OMERO 5.4.6 or newer.

Installing from PyPI

This section assumes that an OMERO.web is already installed.

Install the app using pip:

$ pip install -U omero-parade

Add parade custom app to your installed web apps:

$ bin/omero config append omero.web.apps '"omero_parade"'

Display parade in the centre of the webclient:

$ bin/omero config append omero.web.ui.center_plugins \
    '["Parade", "omero_parade/init.js.html", "omero_parade"]'

Now restart OMERO.web as normal.

Build

In order to build you need:

• npm version equal or greater to 3.0! npm version equal or greater than 5.2 is recommended!

$ npm install

To build an uncompressed version and automatically rebuild when source files change, run:

$ npm run watch

To build an uncompressed version, run:

$ npm run build-dev

To build a compressed, minified version for production, run:

$ npm run build

Custom Filtering

Users can customize the filtering options available by adding their own python modules to the setting:

omero.web.parade.filters

The current default setting lists the omero_parade app itself and two other modules that are in the same directory and are therefore expected to be on the PYTHONPATH when the app is installed.

'["omero_parade", "omero_parade.annotation_filters", "omero_parade.table_filters"]'

Each of these modules contains an omero_filters.py which is expected to implement 2 methods: get_filters and get_script.

The get_filters method is used to compile the list of filters returned by the URL /omero_parade/filters/.

Some examples of get_filters

# Return a list of filter names.
def get_filters(request, conn):
    return ["Rating", "Comment", "Tag"]

The request may include plate or dataset ID if we only want to support the filter for certain data types. In this example we could even check whether an OMERO.table exists on the plate.

def get_filters(request, conn):
    if request.GET.get('plate', None) is not None:
        return ["Table"]
    return []

The get_script function for a named filter should return a JsonResponse that includes a list of parameters for the user input to the filter and a JavaScript filter function.

The JavaScript function will be called for each image to filter and will also be passed in a params object with the user input.

# Return a JS function to filter images by various params.
def get_script(request, script_name, conn):

    dataset_id = request.GET.get('dataset')
    // OR...
    plate_id = request.GET.get('plate')

    if script_name == "Rating":
        # Load rating data for images in Dataset or Wells in Plate...
        # ...
        # var ratings = {imageId: rating} for all images
        var js_object_attr = 'id';  # or 'wellId' if filtering Wells

        # Return a JS function that will be passed an object
        # e.g. {id: 1} for Image or {id: 1, wellId:2} for Image in Well.
        # and should return true or false
        f = """(function filter(data, params) {
            var ratings = %s;
            var match = ratings[data.%s] == params.rating;
            return (params.rating === '-' || match);
        })
        """ % (json.dumps(ratings), js_object_attr)

        filter_params = [{'name': 'rating',
                        'type': 'text',
                        'values': ['-', '1', '2', '3', '4', '5'],
                        'default': '-',
                        }]
        return JsonResponse(
            {
                'f': f,
                'params': filter_params,
            })

Custom Data Providers

Custom data providers return numerical data for Images that can be shown in a table for sorting, or plotted in a graph. NB: Even if data applies to Wells, you need to map this to Image ID, since that is the common denominator that is used to identify images in the various list, grid or plot layouts.

Using the same setup as for filtering above, each module listed in the omero.web.parade.filters setting can also contain a data_providers.py file that implements two methods get_dataproviders and get_data.

Examples for omero_parade/data_providers.py

def get_dataproviders(request, conn):
    return ["ROI_count"]


def get_data(request, data_name, conn):
    """Return data for images in a Dataset or Plate."""
    dataset_id = request.GET.get('dataset')
    plate_id = request.GET.get('plate')
    field_id = request.GET.get('field')

    # ... get img_ids for container, then...

    if data_name == "ROI_count":
        # Want to get ROI count for images
        params = ParametersI()
        params.addIds(img_ids)
        query = "select roi.image.id, count(roi.id) from Roi roi "\
                "where roi.image.id in (:ids) group by roi.image"
        p = query_service.projection(query, params, conn.SERVICE_OPTS)
        roi_counts = {}
        for i in p:
            roi_counts[i[0].val] = i[1].val
        return roi_counts