Eg subtypes of Number would be repacked into a vector with a single
element.

Not a big deal though, I can work around this.

Best,

Tamas

On Fri, May 01 2015, Josh Langsfeld <[email protected]> wrote:

> What's the motivation for wanting to restrict calling repack on
> non-composite types? 'fieldnames' works fine on them too.
>
> If you want to do a different operation on non-composites, wouldn't you
> have to define specialized methods for different type families anyway since
> they would all have different interfaces/behaviors?
>
> On Friday, May 1, 2015 at 10:12:15 AM UTC-4, Tamas Papp wrote:
>
>> The use case is the following: I have parameters to various economic
>> models in composite (immutable) types. Each model maps to a set of
>> moments, for which I have data, and I am calling NLopt to find the
>> parameters for which the moments match the data. For this, I need to be
>> able to convert back and forth between immutable types and vectors
>> (which can be passed to NLopt).
>>
>> I am using the function
>>
>> repack(schema, value, outputtype)
>>
>> The same value may be mapped in multiple ways, hence `schema`. For
>> example, a schema could be that various fields of a composite type are
>> ignored or constrained.
>>
>> A straightforward natural mapping when the schema is the composite type
>> itself is
>>
>> repack(T::DataType, x::T, ::Type{Vector}) = [getfield(x,f) for f =
>> fieldnames(x)]
>>
>> I want to write a short version
>>
>> repack{T}(x::T, outputtype) = repack(T, x, outputtype)
>>
>> that only gets called when T is a DataType. I don't know how to do this.
>>
>> But in the meantime I realized that DataType is not what I want anyway,
>> I want to test for composite types, but that has no explicit type.
>>
>> Best,
>>
>> Tamas
>>
>>
>> On Fri, May 01 2015, Tom Breloff <[email protected] <javascript:>> wrote:
>>
>> > If I understand correctly, you want to pass an instance of Foo into
>> fun2,
>> > which would then effectively call "fun1(Foo, Foo())".  I don't quite
>> > understand the use case... can you give an example of when the second
>> > "fun1" definition would be called?
>> >
>> > On Friday, May 1, 2015 at 9:24:59 AM UTC-4, Tamas Papp wrote:
>> >>
>> >> Not really --- I wanted to dispatch on something being an _instance_ of
>> >> a type that is a DataType. Having reread the manual, I don't think that
>> >> is possible directly, but works with one level of indirection:
>> >>
>> >> fun1(T::DataType,x) = "do something with $x"
>> >> fun1(T, x) = "this is not a datatype"
>> >> fun2{T}(x::T) = fun1(T,x)
>> >> type Foo end
>> >> fun2(Foo())
>> >>
>> >> That said, I realized that DataType is not what I need.
>> >>
>> >> Best,
>> >>
>> >> Tamas
>> >>
>> >> On Thu, Apr 30 2015, Tom Breloff <[email protected] <javascript:>>
>> wrote:
>> >>
>> >> > Is this what you're looking for?
>> >> >
>> >> >
>> >> > julia> yyy(x::DataType) = true
>> >> > yyy (generic function with 1 method)
>> >> >
>> >> >
>> >> > julia> yyy(x) = false
>> >> > yyy (generic function with 2 methods)
>> >> >
>> >> >
>> >> > julia> yyy(Int)
>> >> > true
>> >> >
>> >> >
>> >> > julia> yyy(5)
>> >> > false
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > On Thursday, April 30, 2015 at 11:04:02 AM UTC-4, Tamas Papp wrote:
>> >> >>
>> >> >> This is toy problem that came up in the context of something larger,
>> >> >> reduced to be simple so that I can ask about it more easily.
>> >> >>
>> >> >> Suppose I want to implement the function
>> >> >>
>> >> >> is_instanceof_datatype(x) = is(typeof(x),DataType)
>> >> >>
>> >> >> using dispatch: a default method that is
>> >> >>
>> >> >> is_instanceof_datatype(x) = false
>> >> >>
>> >> >> and some other method which only gets called when x is an instance
>> of a
>> >> >> DataType:
>> >> >>
>> >> >> is_instanceof_datatype{ ... }(x::T) = true # how to dispatch
>> >> >>
>> >> >> but I don't know how to do the latter, hence the ....
>> >> >>
>> >> >> The context is that I want to write a method that, for instances of
>> >> >> DataTypes, returns the slots in a given order, but for other values
>> it
>> >> >> does something else, and I don't know how to do this.
>> >> >>
>> >> >> Best,
>> >> >>
>> >> >> Tamas
>> >> >>
>> >>
>>

Reply via email to