Cannot set linear scaling in certain PhasePlots #3431
Assignees
Labels
Comments
|
Your test script also fails with the current tip of the main branch (7adc079), but with a completely different error. edit: the only difference on main since you reported this is a pre-commit config update, clearly not related. I don't understand why we're getting different errors. |
|
This looks like an unrelated issue and I don't know how it's popping up --
this method should be added in the current main.
…On Tue, Jul 13, 2021 at 11:44 AM Clément Robert ***@***.***> wrote:
Your test script also fails with the current tip of the main branch (
7adc079
<7adc079>),
but with a completely different error.
yt : [INFO ] 2021-07-13 18:41:19,343 Sample dataset found in '/Users/robcleme/dev/yt-project/test-data-dir/FIRE_M12i_ref11/snapshot_600.hdf5'
yt : [INFO ] 2021-07-13 18:41:19,394 Calculating time from 1.000e+00 to be 4.355e+17 seconds
yt : [INFO ] 2021-07-13 18:41:19,395 Assuming length units are in kpc/h (comoving)
yt : [INFO ] 2021-07-13 18:41:19,468 Parameters: current_time = 4.3545571088051405e+17 s
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: domain_dimensions = [1 1 1]
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: domain_left_edge = [0. 0. 0.]
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: domain_right_edge = [60000. 60000. 60000.]
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: cosmological_simulation = 1
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: current_redshift = 0.0
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: omega_lambda = 0.728
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: omega_matter = 0.272
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: omega_radiation = 0.0
yt : [INFO ] 2021-07-13 18:41:19,469 Parameters: hubble_constant = 0.702
yt : [INFO ] 2021-07-13 18:41:19,483 Allocating for 4.787e+06 particles
Initializing coarse index : 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 10/10 [00:00<00:00, 40.88it/s]
╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ /Users/robcleme/dev/yt-project/yt/yt/geometry/particle_geometry_handler.py:181 in │
│ _initialize_index │
│ │
│ 178 │ │ try: │
│ 179 │ │ │ if dont_load: │
│ 180 │ │ │ │ raise OSError │
│ ❱ 181 │ │ │ rflag = self.regions.load_bitmasks(fname) │
│ 182 │ │ │ rflag = self.regions.check_bitmasks() │
│ 183 │ │ │ self._initialize_frontend_specific() │
│ 184 │ │ │ if rflag == 0: │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/geometry/particle_oct_container.pyx:1028 in │
│ yt.geometry.particle_oct_container.ParticleBitmap.load_bitmasks │
│ │
│ 1025 │ │ cdef bytes serial_BAC │
│ 1026 │ │ cdef np.uint64_t ifile │
│ 1027 │ │ f = open(fname,'wb') │
│ ❱ 1028 │ │ # Header │
│ 1029 │ │ f.write(struct.pack('Q', _bitmask_version)) │
│ 1030 │ │ f.write(struct.pack('q', self.file_hash)) │
│ 1031 │ │ f.write(struct.pack('Q', self.nfiles)) │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
OSError
During handling of the above exception, another exception occurred:
╭─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ /Users/robcleme/dev/yt-project/yt/t.py:3 in <module> │
│ │
│ 1 import yt │
│ 2 ds = yt.load_sample('FIRE_M12i_ref11') │
│ ❱ 3 _, c = ds.find_max(('gas', 'density')) │
│ 4 sp = ds.sphere(c, (150.0, 'kpc')) │
│ 5 p = yt.PhasePlot(sp, ('PartType0', 'particle_position_spherical_theta'), ('PartType0', ' │
│ 6 p.set_log(('PartType0', 'particle_position_spherical_theta'), False) │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/data_objects/static_output.py:982 in find_max │
│ │
│ 979 │ │ This is a wrapper around _find_extremum │
│ 980 │ │ """ │
│ 981 │ │ mylog.debug("Searching for maximum value of %s", field) │
│ ❱ 982 │ │ return self._find_extremum(field, "max", source=source, to_array=to_array) │
│ 983 │ │
│ 984 │ def find_min(self, field, source=None, to_array=True): │
│ 985 │ │ """ │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/data_objects/static_output.py:947 in _find_extremum │
│ │
│ 944 │ │ """ │
│ 945 │ │ ext = ext.lower() │
│ 946 │ │ if source is None: │
│ ❱ 947 │ │ │ source = self.all_data() │
│ 948 │ │ method = { │
│ 949 │ │ │ "min": source.quantities.min_location, │
│ 950 │ │ │ "max": source.quantities.max_location, │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/data_objects/static_output.py:1039 in all_data │
│ │
│ 1036 │ │ all_data is a wrapper to the Region object for creating a region │
│ 1037 │ │ which covers the entire simulation domain. │
│ 1038 │ │ """ │
│ ❱ 1039 │ │ self.index │
│ 1040 │ │ if find_max: │
│ 1041 │ │ │ c = self.find_max("density")[1] │
│ 1042 │ │ else: │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/data_objects/static_output.py:526 in index │
│ │
│ 523 │ @Property │
│ 524 │ def index(self): │
│ 525 │ │ if self._instantiated_index is None: │
│ ❱ 526 │ │ │ self._instantiated_index = self._index_class( │
│ 527 │ │ │ │ self, dataset_type=self.dataset_type │
│ 528 │ │ │ ) │
│ 529 │ │ │ # Now we do things that we need an instantiated index for │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/geometry/particle_geometry_handler.py:25 in __init__ │
│ │
│ 22 │ │ self.dataset = weakref.proxy(ds) │
│ 23 │ │ self.float_type = np.float64 │
│ 24 │ │ super().__init__(ds, dataset_type) │
│ ❱ 25 │ │ self._initialize_index() │
│ 26 │ │
│ 27 │ def _setup_geometry(self): │
│ 28 │ │ self.regions = None │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/frontends/sph/data_structures.py:91 in _initialize_index │
│ │
│ 88 │ │ if hasattr(self.io, "_generate_smoothing_length"): │
│ 89 │ │ │ self.io._generate_smoothing_length(self) │
│ 90 │ │ │
│ ❱ 91 │ │ super()._initialize_index() │
│ 92 │ │
│ 93 │ def _generate_kdtree(self, fname): │
│ 94 │ │ from yt.utilities.lib.cykdtree import PyKDTree │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/geometry/particle_geometry_handler.py:188 in │
│ _initialize_index │
│ │
│ 185 │ │ │ │ raise OSError │
│ 186 │ │ except (OSError, struct.error): │
│ 187 │ │ │ self.regions.reset_bitmasks() │
│ ❱ 188 │ │ │ self._initialize_coarse_index() │
│ 189 │ │ │ self._initialize_refined_index() │
│ 190 │ │ │ # We now update fname since index_order2 may have changed │
│ 191 │ │ │ fname = _current_fname() │
│ │
│ /Users/robcleme/dev/yt-project/yt/yt/geometry/particle_geometry_handler.py:230 in │
│ _initialize_coarse_index │
│ │
│ 227 │ │ │ # domain, it will correspond to a very very high index order, which │
│ 228 │ │ │ # is a large amount of memory! Having multiple indexes, one for │
│ 229 │ │ │ # each particle type, would fix this. │
│ ❱ 230 │ │ │ new_order2 = self.regions.update_mi2(max_hsml, ds.index_order[1] + 2) │
│ 231 │ │ │ mylog.info( │
│ 232 │ │ │ │ "Updating index_order2 from %s to %s", ds.index_order[1], new_order2 │
│ 233 │ │ │ ) │
╰──────────────────────────────────────────────────────────────────────────────────────────────────╯
AttributeError: 'yt.geometry.particle_oct_container.ParticleBitmap' object has no attribute 'update_mi2'
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#3431 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAAVXO6AEL5FCQXYC44K4YDTXRUNDANCNFSM5AH23PGA>
.
|
|
@neutrinoceros You need to |
|
Oh that's it, sorry I should have though about it, thank you ! |
meeseeksmachine
pushed a commit
to meeseeksmachine/yt
that referenced
this issue
Oct 15, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Bug report
Bug summary
There appears to be a bug in PhasePlot in how it calculates extrema on subsequent passes, and it leads to failures when trying to changing PhasePlot parameters like which fields are set to log scaling or linear scaling.
Code for reproduction
This code creates a sphere object, then it attempts to create a PhasePlot of spherical theta vs spherical radius.
Actual outcome
The code fails on the line where we try to change the theta field to linear scaling (the second to last line).
Expected outcome
For it to change the theta field to linear scaling and then output a PhasePlot.
Notes
If one removes the second-to-last line:
p.set_log(('PartType0', 'particle_position_spherical_theta'), False), the code snippet above works perfectly fine. In digging into this, the issue appears to in the extrema code here. When yt initially generates the ProfilePlot,extremais equal to None, so it enters the first conditional and correctly calculates that the extrema of the two fields (theta, radius):And everything is fine. But if the last line
p.set_log(('PartType0', 'particle_position_spherical_theta'), False)is included in our script, yt goes back through all of theProfilePlotcode and runs it again (I'm not sure why). This time, though,extremais not None (but it's just composed of all Nones) so it enters the second conditional, which is a bunch of messy code that I cannot really follow. The outcome is that the extrema minima are now negative, and yt fails:I'm submitting a PR that simply checks to see if
extremais composed of all Nones, and if so, it treats it as though it is None, so the same logic occurs. But maybe there is a better way?Version Information
dev tip:
The text was updated successfully, but these errors were encountered: