Message: Re: CLHEP unit definitions Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

None Re: CLHEP unit definitions 

Forum: Documentation and Examples
Re: Question CLHEP unit definitions (Adam Glaser)
Date: 05 Nov, 2012
From: Michael H. Kelsey <Michael H. Kelsey>

On Mon, 05 Nov 2012 13:39:28 GMT, Adam Glaser wrote:
> Does anyone have experience adding unit definitions to CLHEP? I am
> simply trying to add inverse length units which would be very simply
> converted to length units after being inverted. I have added my unit
> definitions, but for some reason the inversion calculation is not being
> performed correctly.

The CLHEP (and GEANT4) unit definitions are nothing more than just floating point (double) scale factors which multiply simple floating point (double) numbers. There are no special "unit aware" classes with overloaded operators.

> i.e., For my inverse cm unit (icm), I am saying that icm=1./cm where
> cm=10*mm. But what is happening is my unit is being converted to mm by
> just multiplying what is entered by 1/10, instead of actually being
> inverted. So GEANT4 is computing 5 icm as 0.5 cm, instead of 0.2 cm.

If you want inverse lengths, then you need to do the inversion of the whole numeric value yourself. Thus:

    G4double myAbsorptionCoeff = 1./(0.5*cm);     // == 2.*icm
    G4double myFrequency = 1./(500.*seconds);     // == 500.*Hz

> Essentially what I am trying to do is analogous to converting hertz to
> seconds. I checked the functionality and it seems that the same problem
> occurs for converted hertz, i.e. it is not properly inverting the value
> entered for hertz to seconds.

Defining a "smart" dimensional class, something named, for example "G4doubleWithUnit", will be very complex. You'll need to defined overloaded operators for arithmetic and math functions, so that the outputs of, e.g., sqrt(), cos(), etc. carry through with the requisite derived units. Such overloading may introduce significant performance issues, since all the math will have to be done with explicit function calls, rather than being optimizable to CPU-based operations.

 Add Message Add Message
to: "Re: CLHEP unit definitions"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews

[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]