Operator

class pyLambdaFlows.op.Operation(parent, funct, topologie, name=None)

The basic pyLambdaElement op.

Operation is a generic operator using a topologie function and a parent operator. This class needs a topologie method to create dependancies with the parent node.

__init__(parent, funct, topologie, name=None)

Initialize self. See help(type(self)) for accurate signature.

compile(sess=None, purge=False)

This function send this lambda op to AWS service.

This function will send it own function to AWS but also all the dependancies one. You have to call this method before the eval one !

eval(feed_dict=None, sess=None, wait=True)

Call this method to compute the result throw the computing graph.

In order to use this function, you have to provide all Source operator default values. To do that, you just have use the kwarg feed_dict as follow :

Example :

a = Source()
b = Map(a)
b.compile()
b.eval(feed_dict={a:[1,2,3]})

class pyLambdaFlows.op.Map(parent, funct, name=None)

The common map operator.

This map operator will apply the given function to all parent elements (one by one), and return the result with the same input size.

__init__(parent, funct, name=None)

Initialize self. See help(type(self)) for accurate signature.

class pyLambdaFlows.op.Reduce(parent, funct, name=None)

The common reduce operator.

This reduce operator will gather all incoming data using the given function. That’s means you hate to provide a function able to deal with a dynamic input size.

You can look at the example mean.

__init__(parent, funct, name=None)

Initialize self. See help(type(self)) for accurate signature.