[yt-dev] Issue #966: Negative radius values in particle_spherical_position_radius fields (and some inconsistant values) (yt_analysis/yt)

[Benjamin Thompson]

New issue 966: Negative radius values in particle_spherical_position_radius fields (and some inconsistant values)
https://bitbucket.org/yt_analysis/yt/issue/966/negative-radius-values-in

Benjamin Thompson:

>From discussion within this PR

[https://bitbucket.org/yt_analysis/yt/pull-request/1308/wip-particle_cylindrical_r-and/diff](https://bitbucket.org/yt_analysis/yt/pull-request/1308/wip-particle_cylindrical_r-and/diff)

I have found that particle_spherical_position_radius (as well as my newly proposed particle_cylindrical_position_radius) produces negative radius values.


```
#!python

sphere["particle_cylindrical_position_radius"].min()
```

-6.41156882532e+23 cm

```
#!python


sphere["particle_spherical_position_radius"].min()
```

-7.83826739716e+23 cm

wheras sphere["particle_spherical_radius"] does not

```
#!python


sphere["particle_spherical_radius"].min()
```

1.25498335796e+20 cm

FYI particle_spherical_radius is more of a direct copy from how the fluid fields compute the spherical radius. 

Interestingly enough. Both these methods produce different results too.

E.g

```
#!python


sphere["particle_spherical_radius"][20]
```
returns 2.37463756333e+23 cm

and

```
#!python

sphere["particle_spherical_position_radius"][20]
```
returns
5.0064012169e+22 cm

FYI the two different methods for calculating these radii are. For "spherical_position_radius"


```
#!python

    def _particle_spherical_position_radius(field, data):
        """
        Radial component of the particles' position vectors in spherical coords
        on the provided field parameters for 'normal', 'center', 
        'bulk_velocity', 
        """
        normal = data.get_field_parameter('normal')
        center = data.get_field_parameter('center')
        bv = data.get_field_parameter("bulk_velocity")
        pos = spos
        pos = YTArray([data[ptype, pos % ax] for ax in "xyz"])
        theta = get_sph_theta(pos, center)
        phi = get_sph_phi(pos, center)
        pos = pos - np.reshape(center, (3, 1))
        sphr = get_sph_r_component(pos, theta, phi, normal)
        return sphr

    registry.add_field((ptype, "particle_spherical_position_radius"),
              function=_particle_spherical_position_radius,
              particle_type=True, units="cm",
              validators=[ValidateParameter("normal"), 
                          ValidateParameter("center")])

```

and "spherical_radius"

  
```
#!python

    def _particle_radius(field, data):
        return get_radius(data, "particle_position_")

    registry.add_field((ptype, "particle_radius"),
              function=_particle_radius,
              validators=[ValidateParameter("center")],
              units="cm", particle_type = True,
              display_name = "Particle Radius")

    # this is just particle radius but add it for ease of use
    registry.add_field((ptype, "particle_spherical_radius"),
              function=_particle_radius,
              particle_type=True, units="cm",
              validators=[ValidateParameter("normal"),
                          ValidateParameter("center")])


```

I have noticed in the fluid gas fields, we are using instead.

```
#!python


    def _spherical_r(field, data):
        coords = get_periodic_rvec(data)
        return data.ds.arr(get_sph_r(coords), "code_length").in_cgs()
```

which returns positive radii as expected.

So basically. Which "radius" calculation should be used for spherical radii? I am thinking a similar sort of manner as done in the fluid fields. 


In summary

1) Why are we getting negative radius for particle_spherical_position_radius 
2) Why



_______________________________________________
yt-dev mailing list
yt-dev at lists.spacepope.org
http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org