add output for environment

currently the profile of the linsley atmosphere that is used in the air shower application is written to earth.dat. This should be done via a writer into a yaml file.

useful information:

• nuclear components
• layers
• density profile

optional: In the IO library access function that lets one use the profiles eg get_grammage(altitude) would be nice to have.

changed milestone to %Beta Release

added Media Output labels

Do you have an idea on this implementation?

These are nested templates. So whereas maybe you would like to do something like GetProperties(summary.yaml), I am not sure that you could construct such a function.

@mreininghaus already made a function to walk through the tree and get all the target materials, so maybe we start from there?

Hi all! I only recently started reading the C8 code and I'm still catching up with the huge amount of work that has gone into it.

An environment output would be very useful also for the Eisvogel integration in !616 (closed) (where it is needed to ensure consistency between the C8 environment and the geometry used for the Green's function).

I had a quick go at a possible implementation. Here is a runnable demonstrator: 48e18eae

Very roughly, every material interface class picks up two new virtual functions ...

template <class T>
class IMaterial : public T {
    
    public:

    // ...
    
    template <typename... Args>
    virtual void DescribeThePropertiesOfThisMaterial(Args&&... args) const = 0;
    
    template <typename... Args>
    virtual void DescribeProperties(Args&&... args) const override {
        T::DescribeProperties(std::forward<Args>(args)...);
        DescribeThePropertiesOfThisMaterial(std::forward<Args>(args)...);
    };
};

... and any concrete medium knows how to describe itself ...

template <class T>
class SomeActualMaterial : public T {
    public:
    
    template <typename... Args>
    void DescribeThePropertiesOfThisMaterial(Args&&... args) const override {
        // ...
    }
};

In an application that defines an environment, calling DescribeProperties will then walk up the inheritance hierarchy and pick up descriptions of all properties of the material.

Of course, all this says nothing about how the materials are actually described ... e.g. in some symbolic way that can be parsed, or as a numerical dump with material properties defined on a discrete grid.

Any thoughts?

Yeah this sounds reasonable. We have made some decision about output that I think we should stick to. So outputting as YAML seems to be the way to go. It is also nice because it is easily parsable by your external software.

Also, I think you are missing the output which describes what the object is e.g. a sphere with center (x,y,z) and radius R.

For your work, this would be equally important. Ideally you can just say universe->dump(yaml) and it walks the whole tree and fills out the yaml with geometric and material properties.

But your work gets us half way there :D

Also, I think you are missing the output which describes what the object is e.g. a sphere with center (x,y,z) and radius R.

For your work, this would be equally important. Ideally you can just say universe->dump(yaml) and it walks the whole tree and fills out the yaml with geometric and material properties.

Indeed! I didn't include the geometry output in the demonstrator because I wanted to hear people's opinions before I go all the way ... but I don't think there are any fundamental challenges there.

We have made some decision about output that I think we should stick to. So outputting as YAML seems to be the way to go. It is also nice because it is easily parsable by your external software.

There are some discussions on Mattermost in parallel to this, on whether we should cook our own schema (YAML or JSON), or whether using something "off-the-shelf" is the option with lower maintenance cost in the long-term (e.g. Geant4 uses GDML, which is XML-based). If you have any thoughts, always happy to hear them!

remove earth.dat for release, come back later

removed milestone %Beta Release

mentioned in issue #708 (closed)

mentioned in merge request corsika-tools!6