ian-r-rose / bayesian_plate_reconstruction Goto Github PK

Make figure with variable trade-off between the two for different priors—joint plot could work well here. Explain in text.

running the make file leads to this exception which is also an issue in invert_one_euler_pole.py and invert_two_euler_poles.py. Is this due to changes made while implementing the TPW inversion?

Traceback (most recent call last):
  File "paleomagnetic_euler_pole.py", line 26, in <module>
    lon_lat = dummy_pole_position_fn(hidden_start_pole, a, hidden_euler_pole, hidden_euler_rate)
  File "/Users/Laurentia/0000_Github/mcplates/mcplates/apw_path.py", line 34, in pole_position
    raise Exception("Unexpected number of euler poles/changpoints: expected %i, got %i"%(n_euler_poles*3-1, len(args)))
Exception: Unexpected number of euler poles/changpoints: expected 2, got 0

I keep getting this UserWarning:

/anaconda3/envs/bayesian_plate/lib/python3.6/site-packages/matplotlib/style/core.py:53: UserWarning: Style includes a parameter, 'savefig.directory', that is not related to style.  Ignoring
  "to style.  Ignoring".format(key))

Is it ok to delete this line from the bpr.mplstyle file?

savefig.directory : 

The modification of the inversion to solve for a great circle should be introduced near the end of this section. Here is the text we have in the Keweenawan Track manuscript:

To solely explain the Keweenawan Track as the result of true polar wander, we can modify the Bayesian inversion to be restricted to inverting the path as a great circle. This inversion is implemented by forcing the Euler pole to be 90\textdegree$;$from the starting point of the path.

It also makes sense to detail that inversion in the Keweenawan example.

While I realize that the figures can be generated via the make file it is computationally expensive to do so. Given that we will need to submit the figure files and it would be nice to see updates graphically (like the one you just processed for Australia) I think it makes sense to have them in there.

The only downside I see is that it makes the .git a bit hefty given it holding on to the versioning of those files. But this doesn't seem like a big deal to me and is something that is fine in other manuscript repos I am working with. Perhaps there are other downsides?