[kepler-dev] Type system.

[supriya gulati]

Hello Ben,

Thank you for your reply.

Output of my actor is displayed using a display actor in the format I
required(i.e a record token) when I do not specify port types. However if I
connect my actor to a record disassembler, there is a type error. Therefore,
I have to set types of the output port as a Record Token where I have
specified the exact type of every element/label of the record. This solves
my problem.

On Mon, Feb 8, 2010 at 7:22 AM, ben leinfelder <leinfelder at nceas.ucsb.edu>wrote:

> Supriya,
> It looks like you've correctly identified the types you're expecting on the
> ports.
> In many cases, the runtime type checking system will 'figure out' the
> correct type for the ports and you won't have to explicitly set them.
> It's non-trivial to define the RecordToken types since they can contain
> wildly heterogenous content.
> Is your workflow running without you specifying the port types?
> -ben
>
>
> On Feb 4, 2010, at 6:22 PM, supriya gulati wrote:
>
>
>> Hello Everyone,
>>
>> I am trying to understand the type system in Kepler in order to construct
>> a typed actor but I am having a hard time. I did refer to the Ptolemy II(Vol
>> II) document for the type system but it looks a little complex.
>>
>> So the problem is that I need to develop an actor which has two input
>> ports named input and window and one output port. Following is the data on
>> all the ports:
>> 1)Input port: {data=x, timestamps=y}
>> 2)Window port:{a,b}
>> 3)output: {Wind={a,b},aggregation={z}}
>>
>> So here Input port and output port has record tokens and window port is an
>> array of two values. Data, timestamps and Wind and aggregation refers to
>> labels in the record type. x is a scalar token, y is a double, a and b are
>> both double, z is scalar. Also the label for aggregation can be min, max,
>> sum etc depending on what value user selects in the parameter field.
>>
>> Can somebody please help me on setting the types for all the ports?
>> --
>> Regards,
>> Supriya
>> _______________________________________________
>> Kepler-dev mailing list
>> Kepler-dev at kepler-project.org
>> http://mercury.nceas.ucsb.edu/kepler/mailman/listinfo/kepler-dev
>>
>
>


-- 
Regards,
Supriya
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mercury.nceas.ucsb.edu/kepler/pipermail/kepler-dev/attachments/20100209/8af54b13/attachment.html>