# You may import any python packages that will be available in the environment you will run this algorithm in

# Environments can change based on the experiment's settings

​

import numpy as np

​

​

class Algorithm:

    # initialise fields to store cross-input data (e.g. machines, aggregations, etc.)

    def __init__(self):

        self.total_der = 0.0

        self.total_time = 0.0

​

    # this will be called each time the sync'd input has more data available to be processed

    def process(self, inputs, dataloader, output):

        # Groups available:

        # Group 0:

        #   Input "DER" with type  "system/float/1"

        #   Input "speech_duration" with type  "system/float/1"

        #   Input "file_info" with type  "allies/file_info/1"

​

        # Results available:

        #   Result "DER" with type "float32"

​

        # to check if there is more data waiting in the inputs

        # (if it is False, you have processed all the inputs and this "process" function won't be called again):

        #       if inputs.hasMoreData():

​

        # to check if a specific input is done:

        #       if inputs["input1"].isDataUnitDone():

​

        # to manually fetch the next input of a specific input

        # (e.g. the block is not sync'd to the input but you want the input immediately)

        #       inputs['input1'].next()

        # you can then access that input value as normal:

        #       self.val1 = inputs['input1'].data

​

        # to get the data for an input (note that the value will be of the type specified in the metadata!):

        #       data_value = inputs['input1'].data

​

        # to write to an output:

        #       outputs['output1'].write({

        #           'output_field_1': 1,

        #           'output_field_2': 'output'

        #       })

​

        # always return True, it signals BEAT to continue processing

        fi =  inputs["file_info"].data.file_id

        sup = inputs["file_info"].data.supervision

        print("file {}: DER = {}, supervision: {}".format(fi, inputs["DER"].data.value, sup))

        tm = inputs["speech_duration"].data.value

        self.total_time += tm

        self.total_der += tm * inputs["DER"].data.value

​

        if not inputs.hasMoreData():

            output.write({"DER": np.cast["float32"](self.total_der/self.total_time)})

​

        return True

​