def build_ui(cls): # define arguments that behave as function inputs inputs = [] inputs.append(UISingleItem(name='dimension_name', datatype=str)) inputs.append( UISingle(name='dimension_value', datatype=str, description='Dimension Filter Value')) inputs.append( UIExpression( name='expression', description="Define alert expression using pandas systax. \ Example: df['inlet_temperature']>50. ${pressure} will be substituted \ with df['pressure'] before evaluation, ${} with df[<dimension_name>]" )) inputs.append( UISingle(name='pulse_trigger', description= "If true only generate alerts on crossing the threshold", datatype=bool)) # define arguments that behave as function outputs outputs = [] outputs.append( UIFunctionOutSingle(name='alert_name', datatype=bool, description='Output of alert function')) outputs.append( UIFunctionOutSingle(name='alert_end', datatype=dt.datetime, description='End of pulse triggered alert')) return (inputs, outputs)
def build_ui(cls): ''' Preload function has no dataframe in or out so standard _getMetadata() does not work ''' # define arguments that behave as function inputs inputs = {} inputs['dummy'] = UISingle( name='dummy', datatype=str, description='Dummy attribute as input').to_metadata() # define arguments that behave as function outputs outputs = OrderedDict() outputs['predict1'] = UIFunctionOutSingle( name='predict1', datatype=str, description='Returns a prediction value').to_metadata() outputs['predict2'] = UIFunctionOutSingle( name='predict2', datatype=str, description='Returns a prediction value').to_metadata() outputs['predict3'] = UIFunctionOutSingle( name='predict3', datatype=str, description='Returns a prediction value').to_metadata() outputs['predict4'] = UIFunctionOutSingle( name='predict4', datatype=str, description='Returns a prediction value').to_metadata() return (inputs, outputs)
def build_ui(cls): # define arguments that behave as function inputs inputs = [] inputs.append(UISingleItem( name='input_item1', datatype=str, description='String encoded array of sensor readings' )) inputs.append(UISingleItem( name='input_item2', datatype=str, description='String encoded array of sensor readings' )) inputs.append(UISingleItem( name='input_item3', datatype=str, description='String encoded array of sensor readings' )) # define arguments that behave as function outputs outputs = [] outputs.append(UIFunctionOutSingle( name='output_item', datatype=float, description='L2 norm of the string encoded sensor readings' )) return (inputs, outputs)
def build_ui(cls): inputs = [] inputs.append( UISingleItem(name='input_item', datatype=float, description='Item to base anomaly on')) inputs.append( UISingle( name='factor', datatype=int, description= 'Frequency of anomaly e.g. A value of 3 will create anomaly every 3 datapoints', default=10)) inputs.append( UISingle(name='width', datatype=int, description='Width of the anomaly created', default=5)) outputs = [] outputs.append( UIFunctionOutSingle( name='output_item', datatype=float, description='Generated Item With Flatline anomalies')) return (inputs, outputs)
def build_ui(cls): inputs = [] inputs.append( UISingleItem(name='input_item', datatype=float, description='Item to base anomaly on')) inputs.append( UISingle( name='factor', datatype=int, description= 'Frequency of anomaly e.g. A value of 3 will create anomaly every 3 datapoints', default=5)) inputs.append( UISingle( name='size', datatype=int, description= 'Size of extreme anomalies to be created. e.g. 10 will create 10x size extreme \ anomaly compared to the normal variance', default=10)) outputs = [] outputs.append( UIFunctionOutSingle( name='output_item', datatype=float, description='Generated Item With Extreme anomalies')) return (inputs, outputs)
def build_ui(cls): # define arguments that behave as function inputs inputs = [] inputs.append(UIExpression(name='expression_constant', description="Define alert expression to load as constant using pandas systax. \ Example: df['inlet_temperature']>50. ${pressure} will be substituted \ with df['pressure'] before evaluation, ${} with df[<dimension_name>]")) # define arguments that behave as function outputs outputs = [] outputs.append(UIFunctionOutSingle(name='alert_name', datatype=bool, description='Output of alert function')) return (inputs, outputs)
def build_ui(cls): # define arguments that behave as function inputs inputs = [] inputs.append( UIExpression( name='conditional_expression', description="expression that returns a True/False value, \ eg. if df['temp']>50 then df['temp'] else None" )) inputs.append( UIExpression(name='true_expression', description="expression when true, eg. df['temp']")) inputs.append( UIExpression(name='false_expression', description='expression when false, eg. None')) # define arguments that behave as function outputs outputs = [] outputs.append( UIFunctionOutSingle(name='output_item', datatype=bool, description='Dummy function output')) return (inputs, outputs)