Added 3D Plotting Functionality to plotter.py #584
Conversation
plotter3.py is an additional plotter function that functions similarly to plotter.py, though is meant to visualize targets and measurements in 3D as opposed to 2D planes. Currently, only truth visualization is supported.
Can now plot both measurements and tracks in 3D, functionality & use is identical to plotter.py to maintain consistency in implementation. Todo - add error bars to legend, consider alternate ways to display track uncertainty
-*NEW* Added 3D error bars as option for plot_tracks()
-*NEW* err_freq is an attribute of plot_tracks() that defines how frequently error bars should be displayed along the track (default is 1, or for every point)
-labels_list and handles_list removed, replaced by legend_dict.
--legend_dict holds *_handle with
*_label as the key
-legend is now plotted with legend_dict.values() as the handles and legend_dict.keys() as the labels; avoids duplicate entries in legend if plotting the same data multiple times. If similar elements are desired to be displayed in legend, a unique *_label must be passed into the appropriate plot function
*TODO*
-Redo header documentation for Plotter class and methods
-Add buttons to have interactive plots (turning on and off measurements, tracks, truth, etc.). See https://matplotlib.org/stable/gallery/widgets/check_buttons.html
-Allow user to plot tracks and truths with unique legend identifiers WITHOUT having to call plot_tracks() for each unique element; list of labels same length as list of tracks/truth
Using flake8 python library
plotter.py now contains all functionality of original 2D plotter.py and 3D plotting of plotter3.py. Plotting type is now chosen by the user by using the optional Enumerate parameter "dimension" when instantiating Plotter(). Documentation has been updated to reflect these changes TODO -Add unit tests -Use flake8 to ensure the file complies with PEP8 standards -Submit pull request on Stone Soup Github repo
Deprecated - functionality now contained fully within plotter.py
Delete cached python libraries
PACarniglia
requested review from
sdhiscocks and
svidal-dstl
and removed request for
a team
Codecov Report
@@ Coverage Diff @@
## main #584 +/- ##
==========================================
- Coverage 94.22% 93.95% -0.27%
==========================================
Files 144 144
Lines 7008 7035 +27
Branches 1037 1336 +299
==========================================
+ Hits 6603 6610 +7
- Misses 307 325 +18
- Partials 98 100 +2
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
Co-authored-by: Nikki Perree <[email protected]>
Co-authored-by: Nikki Perree <[email protected]>
Co-authored-by: Nikki Perree <[email protected]>
Co-authored-by: Nikki Perree <[email protected]>
As per @sdhiscocks recommendations
Should rectify tests 36/37
|
@sdhiscocks I was told by @ekhunter123 that the remaining failing tests can be ignored, is this still the case for this PR? |
I think here I was referencing the comment on my pull request #587, that CircleCi is having a problem lately with the dependency cache. I told @PACarniglia about this and wondered if the same applies for this PR. Perhaps "ignore" was too strong of a word for me to use. The fact that the tests fail still gives important information and maybe more work can be done to fix it. On the other hand, maybe it is just a problem with CircleCi, I don't know. |
|
You are correct @ekhunter123, the tests are broken for pull requests; the same dependency cache issue. I'll see if I can work out a way to resolve this, but any fix will probably only resolve itself for new PRs. |
There was a problem hiding this comment.
Looks good 👍
Minor suggestion:
diff --git a/stonesoup/plotter.py b/stonesoup/plotter.py
index 9cdc0c80..999f7566 100644
--- a/stonesoup/plotter.py
+++ b/stonesoup/plotter.py
@@ -53,11 +53,7 @@ class Plotter:
"""
def __init__(self, dimension=Dimension.TWO):
- if isinstance(dimension, type(Dimension.TWO)):
- self.dimension = dimension
- else:
- raise TypeError("""%s is an unsupported type for \'dimension\';
- expected type %s""" % (type(dimension), type(Dimension.TWO)))
+ self.dimension = Dimension(dimension)
# Generate plot axes
self.fig = plt.figure(figsize=(10, 6))
if self.dimension is Dimension.TWO: # 2D axes
diff --git a/stonesoup/tests/test_plotter.py b/stonesoup/tests/test_plotter.py
index 258dab64..46cd7ed2 100644
--- a/stonesoup/tests/test_plotter.py
+++ b/stonesoup/tests/test_plotter.py
@@ -81,7 +81,7 @@ plotter = Plotter()
def test_dimension_raise():
- with pytest.raises(TypeError):
+ with pytest.raises(ValueError):
Plotter(dimension=1) # expected to raise TypeError
Wrapping the input with Enum should allow both Plotter(dimension=3) and Plotter(dimension=Dimension.THREE) to work.
@sdhiscocks the reason for switching the current implementation of dimension as an int to dimension as an Enum is twofold; it sanitizes the input as users will have to use a Dimension instance (this was recommended by @ekhunter123), and makes it easier for future developers to extend the plotting capabilities present in Plotter.py without having to be restricted to just adding additional dimensions - in theory, the dimension parameter need not be a spatial number but could rather be sort of a "plot type" attribute. |
Adds three-dimensional plotting functionality to the plotter.py function. Plotting of truth data, measurements (including clutter and true detection distinction), and tracks with error bars are supported. The underlying 2D plotting functionality is unchanged. Plotting dimensionality is now selected by the user when instantiating new Plotter objects by passing a plotter.py Dimension Enum object. Currently Dimension.TWO and Dimension.THREE are supported but can be extended in the future.
The legend is now populated with a dictionary where the keys show up as labels and values associated with those keys appear as the handles. This implementation avoids duplicate legend entries if plotting the same data multiple times.
A unit test file, test_plotter.py, was also created to ensure subsequent changes to plotter.py do not break current functionality.