issue-250-ridge-transforms

[Hojat-Shirmard]

Search for "hjt" in the file; you'll see I've added lines of code in four different places.

[michaelchin]

@Hojat-Shirmard

• I have changed the title to "issue-250-ridge-transforms". It is more descriptive.
• Can you move the two functions to the end of the file and fix the indentation?
• let's add "ridge_transforms" as property(return self.ridges + self.transforms), similar to this

  gplately/gplately/plot.py

  Lines 383 to 387 in bea66aa

  	@property
  	def topological_plate_boundaries(self):
  	if self._topological_plate_boundaries is None:
  	self.update_time(self.time)
  	return self._topological_plate_boundaries

• create a test program, similar to https://github.com/GPlates/gplately/blob/issue-250-ridge-transforms/tests-dir/unittest/test_plot_with_raster.py (remove the parts which you don't need.)

[Hojat-Shirmard]

Hi @michaelchin
I made a second commit considering your requested changes. To test the code, should I add plot.plot.ridge_transforms to the code you shared and run it in the Jupyter notebook?

[michaelchin]

Hi @michaelchin I made a second commit considering your requested changes. To test the code, should I add plot.plot.ridge_transforms to the code you shared and run it in the Jupyter notebook?

yes, add plot_ridge_transform() funciton call and comment out other plotting code to see the result more clearly.

And you can see your code changes here https://github.com/GPlates/gplately/pull/298/files. there are still errors in the code changes.

[Hojat-Shirmard]

How can I know where the errors are?

[michaelchin]

How can I know where the errors are?

Have you tried to click this link https://github.com/GPlates/gplately/pull/298/files?

if you did, see the screenshot below for example.

You need to make sure the changes are intended. And only make changes when it is necessary.

[Hojat-Shirmard]

I reckon I need to create a new environment in Anaconda and install my modified branch to check if it’s working properly, right?
I've also passed some courses on YouTube and learned a bit about Docker, so if there are any instructions or steps to follow with Docker, I’d be happy to try that as well.

[Hojat-Shirmard]

I added logger.debug
I added the "ridge_transforms" attribute, but the result is not correct!

[jcannon-gplates]

Maybe it's because the @property decorator is missing. See Michael's example above...

 @property 
 def topological_plate_boundaries(self): 
     if self._topological_plate_boundaries is None: 
         self.update_time(self.time) 
     return self._topological_plate_boundaries

You can see here how properties are added.

[Hojat-Shirmard]

Hi @jcannon-gplates
Thank you for your quick response!
I just fixed the property, and here is the result:

[jcannon-gplates]

It looks like you're using self._topological_plate_boundaries. That will work since it'll generate self.ridges and self.transforms if those attributes don't exist yet.

However, it might be better to do something like...

    @property
    def ridge_transforms(self):
        """Combine ridges and transforms into one feature set."""
        if self.time is None:
            raise ValueError(
                "No topologies have been resolved. Set `PlotTopologies.time` to construct them."
            )
        logger.debug("Returning ridge_transforms property.")
        return self.ridges + self.transforms

...where I've replaced...

        if self._topological_plate_boundaries is None:
            logger.debug("Accessing ridge transforms - updating topological plate boundaries.")
            self.update_time(self.time)

...with...

        if self.time is None:
            raise ValueError(
                "No topologies have been resolved. Set `PlotTopologies.time` to construct them."
            )

...because if the time hasn't been set (in either __init__ or with self.time = ...), and hence self.time (as well as self._topological_plate_boundaries) is None, then self.update_time(self.time) will fail here.

@GPlates/gplately-dev, does this sound right? I think self.update_time() probably should only be called on setters (like self.anchor_plate_id = ...). Getters should already be up-to-date (since update_time() has already been called behind the scenes whenever the class state changed, via setters, etc). Getters (that rely on variables updated by update_time()) should perhaps instead raise an error (if the time has not been set yet). My understanding is the user should eventually set the time if they want to access certain functionalities of the class. For example, self.ridges and self.transforms don't exist if the user has not yet set the time, so an error should be raised if they try to access them.

[Hojat-Shirmard]

Thank you so much! I just applied your comment.

[michaelchin]

Thanks @jcannon-gplates and @Hojat-Shirmard

Just out of curiosity, what was the reason of these two lines got removed?

[Hojat-Shirmard]

I think I messed up—sorry about that! I'll fix it now.

[michaelchin]

@Hojat-Shirmard Can you add other properties and functions which were mentioned in #250? And also add the warning and debug messages mentioned in #250. Thanks.

[Hojat-Shirmard]

@michaelchin, Did you mean this comment?
#250 (comment)

[michaelchin]

@michaelchin, Did you mean this comment? #250 (comment)

Could you please read #250? There are a few comments regarding what need to be done. Thanks.

[Hojat-Shirmard]

I added misc_transforms