Patrick,
I find Andrew Frank's work on axiomatic specifications of GIS systems
-- which the paper you cite is built on -- very confusing, or indeed,
confused. They have a bunch of example like
data Car = Car Color
class Car a where
carColor :: a - Color
instance Car Car where
carColor
On 29/01/2011 20:56, Henning Thielemann wrote:
Is there a reason why you use an individual type for every unit?
The existing implementations of typed physical units only encode the
physical dimension in types and leave the unit factors to the value
level. I found this to be the most natural
Patrick Browne schrieb:
On 29/01/2011 20:56, Henning Thielemann wrote:
Is there a reason why you use an individual type for every unit?
The existing implementations of typed physical units only encode the
physical dimension in types and leave the unit factors to the value
level. I found this
On 30/01/2011 19:43, Henning Thielemann wrote:
I do not see a constant 1 that is equated with a type.
This is due to my misunderstanding of Haskell.
After your comments my understanding of the unit function is as follows:
1) In the instance below the argument for unit must have type
Patrick Browne schrieb:
On 30/01/2011 19:43, Henning Thielemann wrote:
I do not see a constant 1 that is equated with a type.
This is due to my misunderstanding of Haskell.
After your comments my understanding of the unit function is as follows:
1) In the instance below the argument for
Hi,
Your definition of 'unit' in the
instance MetricDescription LengthInCentimetres Centimetre
is not well-typed. Maybe you want to write either
unit (LengthInCentimitres 2.0) = Centimetre
-- (pattern match fail for all (LengthInCentimetres l), l /= 2.0)
or
unit l = Centimetre
-- i.e. unit
Patrick Browne schrieb:
Below is some code that is produces information about the *types* used
for measuring (e.g. metres). The following evaluation returns 1.00
which the convert factor for metres.
convertFactorToBaseUnit (unit (LengthInMetres 7))
.
The next evaluation returns the type,