On Fri, May 17, 2013 at 3:39 PM, David Collins wrote:

Is there an easy way to turn data._get_cut_mask(grid) into an AMR region?

Then I could just do a boolean "or" on those regions from each of the regions?

Oh, you want to do it grid-by-grid? I assumed that you were operating on the full clump. If you have the full clump, you can feed that (which should be an AMRExtractedRegion) to the boolean object. And now that I think about it, AMRExtractedRegion is what should have the _flush_to_grids method.

d.

On Fri, May 17, 2013 at 1:34 PM, Matthew Turk wrote:

...

On Fri, May 17, 2013 at 3:31 PM, david collins wrote:

...

Hi, all--

I have a clump that I've gotten with the clump finder. I'd like to paint those zones of that clump with a flag, so I can exclude that region from other analysis. I'm a little stuck on how to get this done-- I feel like it needs to involve clump.data._flush_data_to_grids and a cut_mask, but I'm a little stuck on how to define the field that I flush_to_grids.

I have to think about specifically how you might do this, but I do believe that you can probably accomplish what you're looking for by using a boolean region.

...

Specifically, how to get the cut_mask information to the field, and how to avoid clobbering the paint from one clump with another (or, from one grid with another)

I'm trying to do something like:

<code> for grid in clump.data._grids: cut_mask = clump.data._get_cut_mask(grid) def masked_region(field,data): cut_mask = data.get_field_parameter('cut_mask', None) try: base = data['masked_region'] except: base = na.zeros(data['Density'].shape) if cut_mask is not None: base = na.maximum(base,cut_mask) return base

add_field('masked_region',function=masked_region,validators=[ValidateGridType()], not_in_all=True) grid.set_field_parameter('cut_mask',cut_mask) print grid, cut_mask.sum(), grid['masked_region'].sum() clump.data._flush_to_grids('masked_region',0) </code>

but clump['masked_region'].sum() only returns the flagged zones in the first of the grids.

I feel like this is something I've done before, but can't find any reference to it in my email or old scripts. Apologies if this has been answered elsewhere, or it's obvious and I'm being thick.

Thanks! d.

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

I believe that is correct, yes. One very simple way to test this and verify would be to compare the volume in that object plus the volume in your clump against the volume in the domain. On Fri, May 17, 2013 at 3:48 PM, David Collins wrote:

Oh, right. I can do

pf.h.boolean([pf.h.all_data(), "NOT", clump])

?

On Fri, May 17, 2013 at 1:41 PM, Matthew Turk wrote:

...

On Fri, May 17, 2013 at 3:39 PM, David Collins wrote:

...

Is there an easy way to turn data._get_cut_mask(grid) into an AMR region?

Then I could just do a boolean "or" on those regions from each of the regions?

Oh, you want to do it grid-by-grid? I assumed that you were operating on the full clump. If you have the full clump, you can feed that (which should be an AMRExtractedRegion) to the boolean object.

And now that I think about it, AMRExtractedRegion is what should have the _flush_to_grids method.

...

d.

On Fri, May 17, 2013 at 1:34 PM, Matthew Turk wrote:

...

On Fri, May 17, 2013 at 3:31 PM, david collins wrote:

...

Hi, all--

I have a clump that I've gotten with the clump finder. I'd like to paint those zones of that clump with a flag, so I can exclude that region from other analysis. I'm a little stuck on how to get this done-- I feel like it needs to involve clump.data._flush_data_to_grids and a cut_mask, but I'm a little stuck on how to define the field that I flush_to_grids.

I have to think about specifically how you might do this, but I do believe that you can probably accomplish what you're looking for by using a boolean region.

...

Specifically, how to get the cut_mask information to the field, and how to avoid clobbering the paint from one clump with another (or, from one grid with another)

I'm trying to do something like:

<code> for grid in clump.data._grids: cut_mask = clump.data._get_cut_mask(grid) def masked_region(field,data): cut_mask = data.get_field_parameter('cut_mask', None) try: base = data['masked_region'] except: base = na.zeros(data['Density'].shape) if cut_mask is not None: base = na.maximum(base,cut_mask) return base

add_field('masked_region',function=masked_region,validators=[ValidateGridType()], not_in_all=True) grid.set_field_parameter('cut_mask',cut_mask) print grid, cut_mask.sum(), grid['masked_region'].sum() clump.data._flush_to_grids('masked_region',0) </code>

but clump['masked_region'].sum() only returns the flagged zones in the first of the grids.

I feel like this is something I've done before, but can't find any reference to it in my email or old scripts. Apologies if this has been answered elsewhere, or it's obvious and I'm being thick.

Thanks! d.

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

This is a bug; I have no idea why we're putting arguments into __repr__ like this, but it can be fixed by updating the __repr__ definition in extracted region base. Can you submit a PR? On Fri, May 17, 2013 at 4:07 PM, David Collins wrote:

Ah, Sweet. I tried this:

...

...

...

pf.h.boolean([pf.h.all_data(), 'NOT', clump.data])

and got this:

Building boolean100% |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| Time: 00:00:01 Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/nics/b/home/collins/local_2013_03_12/src/yt-hg/yt/data_objects/data_containers.py", line 4104, in __repr__ s += region.__repr__(clean = True) TypeError: wrapper __repr__ doesn't take keyword arguments

c32.data is of type ExtractedRegion. It works ok if I use pf.h.sphere. What do I need to do to condition the clump.data?

d.

On Fri, May 17, 2013 at 1:49 PM, Matthew Turk wrote:

...

I believe that is correct, yes. One very simple way to test this and verify would be to compare the volume in that object plus the volume in your clump against the volume in the domain.

On Fri, May 17, 2013 at 3:48 PM, David Collins wrote:

...

Oh, right. I can do

pf.h.boolean([pf.h.all_data(), "NOT", clump])

?

On Fri, May 17, 2013 at 1:41 PM, Matthew Turk wrote:

...

On Fri, May 17, 2013 at 3:39 PM, David Collins wrote:

...

Is there an easy way to turn data._get_cut_mask(grid) into an AMR region?

Then I could just do a boolean "or" on those regions from each of the regions?

Oh, you want to do it grid-by-grid? I assumed that you were operating on the full clump. If you have the full clump, you can feed that (which should be an AMRExtractedRegion) to the boolean object.

And now that I think about it, AMRExtractedRegion is what should have the _flush_to_grids method.

...

d.

On Fri, May 17, 2013 at 1:34 PM, Matthew Turk wrote:

...

On Fri, May 17, 2013 at 3:31 PM, david collins wrote: > Hi, all-- > > I have a clump that I've gotten with the clump finder. I'd like > to > paint > those zones of that clump with a flag, so I can exclude that > region > from > other analysis. I'm a little stuck on how to get this done-- I > feel > like it > needs to involve clump.data._flush_data_to_grids and a cut_mask, > but > I'm > a > little stuck on how to define the field that I flush_to_grids. >

I have to think about specifically how you might do this, but I do believe that you can probably accomplish what you're looking for by using a boolean region.

> Specifically, how to get the cut_mask information to the field, > and > how > to > avoid clobbering the paint from one clump with another (or, from > one > grid > with another) > > I'm trying to do something like: > > <code> > for grid in clump.data._grids: > cut_mask = clump.data._get_cut_mask(grid) > def masked_region(field,data): > cut_mask = data.get_field_parameter('cut_mask', None) > try: > base = data['masked_region'] > except: > base = na.zeros(data['Density'].shape) > if cut_mask is not None: > base = na.maximum(base,cut_mask) > return base > > > > > add_field('masked_region',function=masked_region,validators=[ValidateGridType()], > not_in_all=True) > grid.set_field_parameter('cut_mask',cut_mask) > print grid, cut_mask.sum(), grid['masked_region'].sum() > clump.data._flush_to_grids('masked_region',0) > </code> > > but clump['masked_region'].sum() only returns the flagged zones in > the > first > of the grids. > > I feel like this is something I've done before, but can't find any > reference > to it in my email or old scripts. Apologies if this has been > answered > elsewhere, or it's obvious and I'm being thick. > > Thanks! > d. > > > > _______________________________________________ > yt-users mailing list > yt-users@lists.spacepope.org > http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org > _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

I can update that PR to check for cases in which __repr__ accepts that keyword. What kind of object would that be? On Fri, May 17, 2013 at 2:45 PM, Stephen Skory wrote:

...

This is a bug; I have no idea why we're putting arguments into __repr__ like this, but it can be fixed by updating the __repr__ definition in extracted region base. Can you submit a PR?

I added this to make the __repr__ of boolean regions more readable, I think. But if this is breaking things, by all means, remove it!

-- Stephen Skory s@skory.us http://stephenskory.com/ 510.621.3687 (google voice) _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org