Avoid import star (from open3d import *) (isl-org#982)

* import open3d as o3d

* rename python namespace

* format python code with yapf

* add python style

* add init.py

* fix import *

* update all line numbers

* address review comments

* check-style and apply-style for python

.style.yapf

@@ -0,0 +1,2 @@
+[style]
+based_on_style = google

.travis.yml

@@ -32,6 +32,7 @@ matrix:
         - mkdir build
         - cd build
         - cmake -DPYTHON_EXECUTABLE=`which python` ..
+        - pip install -U yapf
         - make check-style
 
     # Build docs only

docs/tutorial/Advanced/color_map_optimization.rst

@@ -18,7 +18,7 @@ Input
 .. literalinclude:: ../../../examples/Python/Advanced/color_map_optimization.py
    :language: python
    :lineno-start: 19
-   :lines: 19-35
+   :lines: 19-38
    :linenos:
 
 This script reads color and depth image pairs and makes ``rgbd_image``. Note that ``convert_rgb_to_intensity`` flag is ``False``. This is to preserve 8-bit color channels instead of using single channel float type image.
@@ -27,16 +27,16 @@ It is always good practice to visualize RGBD image before applying it to color m
 
 .. literalinclude:: ../../../examples/Python/Advanced/color_map_optimization.py
    :language: python
-   :lineno-start: 37
-   :lines: 37-39
+   :lineno-start: 40
+   :lines: 40-44
    :linenos:
 
 The script reads camera trajectory and mesh.
 
 .. literalinclude:: ../../../examples/Python/Advanced/color_map_optimization.py
    :language: python
-   :lineno-start: 43
-   :lines: 43-48
+   :lineno-start: 46
+   :lines: 46-53
    :linenos:
 
 To visualize how the camera poses are not good for color mapping, this script intentionally set the iteration number as 0, which means no optimization. ``color_map_optimization`` paints a mesh using corresponding RGBD images and camera poses. Without optimization, the texture map is blurred.
@@ -55,7 +55,7 @@ The next step is to optimize camera poses to get a sharp color map.
 .. literalinclude:: ../../../examples/Python/Advanced/color_map_optimization.py
    :language: python
    :lineno-start: 55
-   :lines: 55-60
+   :lines: 55-65
    :linenos:
 
 The script sets ``maximum_iteration = 300`` for actual iterations. The optimization displays the following energy profile.

docs/tutorial/Advanced/colored_pointcloud_registration.rst

@@ -50,7 +50,7 @@ Point-to-plane ICP
 .. literalinclude:: ../../../examples/Python/Advanced/colored_pointcloud_registration.py
    :language: python
    :lineno-start: 29
-   :lines: 29-37
+   :lines: 29-38
    :linenos:
 
 We first run :ref:`point_to_plane_icp` as a baseline approach. The visualization below shows misaligned green triangle textures. This is because geometric constraint does not prevent two planar surfaces from slipping.
@@ -89,8 +89,8 @@ To further improve efficiency, [Park2017]_ proposes a multi-scale registration s
 
 .. literalinclude:: ../../../examples/Python/Advanced/colored_pointcloud_registration.py
    :language: python
-   :lineno-start: 39
-   :lines: 39-70
+   :lineno-start: 40
+   :lines: 40-74
    :linenos:
 
 In total, 3 layers of multi-resolution point clouds are created with :ref:`voxel_downsampling`. Normals are computed with :ref:`vertex_normal_estimation`. The core registration function ``registration_colored_icp`` is called for each layer, from coarse to fine.  ``lambda_geometric`` is an optional argument for ``registration_colored_icp`` that determines :math:`\lambda \in [0,1]` in the overall energy :math:`\lambda E_{G} + (1-\lambda) E_{C}`.

docs/tutorial/Advanced/customized_visualization.rst

@@ -13,8 +13,8 @@ Mimic draw_geometries() with Visualizer class
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 12
-   :lines: 12-19
+   :lineno-start: 13
+   :lines: 13-20
    :linenos:
 
 This function produces exactly the same functionality of the convenient function ``draw_geometries``.
@@ -26,8 +26,8 @@ Class ``Visualizer`` has a couple of variables such as a ``ViewControl`` and a `
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 39
-   :lines: 39-46
+   :lineno-start: 46
+   :lines: 46-52
    :linenos:
 
 Outputs:
@@ -42,8 +42,8 @@ To change field of view of the camera, it is necessary to get an instance of vis
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 21
-   :lines: 21-30
+   :lineno-start: 23
+   :lines: 23-32
    :linenos:
 
 The field of view can be set as [5,90] degree. Note that ``change_field_of_view`` adds specified FoV on the current FoV. By default, visualizer has 60 degrees of FoV. Calling the following code
@@ -74,8 +74,8 @@ Use callback functions
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 32
-   :lines: 32-37
+   :lineno-start: 35
+   :lines: 35-43
    :linenos:
 
 Function ``draw_geometries_with_animation_callback`` registers a Python callback function ``rotate_view`` as the idle function of the main loop. It rotates the view along the x-axis whenever the visualizer is idle. This defines an animation behavior.
@@ -85,8 +85,8 @@ Function ``draw_geometries_with_animation_callback`` registers a Python callback
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 48
-   :lines: 48-72
+   :lineno-start: 55
+   :lines: 55-84
    :linenos:
 
 Callback functions can also be registered upon key press event. This script registered four keys. For example, pressing :kbd:`k` changes the background color to black.
@@ -99,8 +99,8 @@ Capture images in a customized animation
 
 .. literalinclude:: ../../../examples/Python/Advanced/customized_visualization.py
    :language: python
-   :lineno-start: 74
-   :lines: 74-118
+   :lineno-start: 87
+   :lines: 87-134
    :linenos:
 
 This function reads a camera trajectory, then defines an animation function ``move_forward`` to travel through the camera trajectory. In this animation function, both color image and depth image are captured using ``Visualizer.capture_depth_float_buffer`` and ``Visualizer.capture_screen_float_buffer`` respectively. They are saved in files.

docs/tutorial/Advanced/fast_global_registration.rst

@@ -20,8 +20,8 @@ Input
 
 .. literalinclude:: ../../../examples/Python/Advanced/fast_global_registration.py
    :language: python
-   :lineno-start: 27
-   :lines: 27-29
+   :lineno-start: 29
+   :lines: 29-31
    :linenos:
 
 For the pair comparison, the script reuses the ``prepare_dataset`` function defined in :ref:`global_registration`.
@@ -32,8 +32,8 @@ Baseline
 
 .. literalinclude:: ../../../examples/Python/Advanced/fast_global_registration.py
    :language: python
-   :lineno-start: 31
-   :lines: 31-37
+   :lineno-start: 33
+   :lines: 33-40
    :linenos:
 
 This script calls RANSAC based :ref:`global_registration` as a baseline. After registration it displays the following result.
@@ -52,8 +52,8 @@ With the same input used for a baseline, the next script calls the implementatio
 
 .. literalinclude:: ../../../examples/Python/Advanced/fast_global_registration.py
    :language: python
-   :lineno-start: 14
-   :lines: 14-22
+   :lineno-start: 15
+   :lines: 15-24
    :linenos:
 
 This script displays the following result.

docs/tutorial/Advanced/global_registration.rst

@@ -16,8 +16,8 @@ Input
 
 .. literalinclude:: ../../../examples/Python/Advanced/global_registration.py
    :language: python
-   :lineno-start: 34
-   :lines: 34-43
+   :lineno-start: 39
+   :lines: 39-50
    :linenos:
 
 This script reads a source point cloud and a target point cloud from two files. They are misaligned with an identity matrix as transformation.
@@ -32,8 +32,8 @@ Extract geometric feature
 
 .. literalinclude:: ../../../examples/Python/Advanced/global_registration.py
    :language: python
-   :lineno-start: 19
-   :lines: 19-31
+   :lineno-start: 21
+   :lines: 21-36
    :linenos:
 
 We down sample the point cloud, estimate normals, then compute a FPFH feature for each point. The FPFH feature is a 33-dimensional vector that describes the local geometric property of a point. A nearest neighbor query in the 33-dimensinal space can return points with similar local geometric structures. See [Rasu2009]_ for details.
@@ -45,8 +45,8 @@ RANSAC
 
 .. literalinclude:: ../../../examples/Python/Advanced/global_registration.py
    :language: python
-   :lineno-start: 49
-   :lines: 49-61
+   :lineno-start: 53
+   :lines: 53-66
    :linenos:
 
 We use RANSAC for global registration. In each RANSAC iteration, ``ransac_n`` random points are picked from the source point cloud. Their corresponding points in the target point cloud are detected by querying the nearest neighbor in the 33-dimensional FPFH feature space. A pruning step takes fast pruning algorithms  to quickly reject false matches early.
@@ -75,8 +75,8 @@ For performance reason, the global registration is only performed on a heavily d
 
 .. literalinclude:: ../../../examples/Python/Advanced/global_registration.py
    :language: python
-   :lineno-start: 64
-   :lines: 64-71
+   :lineno-start: 69
+   :lines: 69-77
    :linenos:
 
 Outputs a tight alignment. This summarizes a complete pairwise registration workflow.

docs/tutorial/Advanced/interactive_visualization.rst

@@ -20,8 +20,8 @@ Crop geometry
 
 .. literalinclude:: ../../../examples/Python/Advanced/interactive_visualization.py
    :language: python
-   :lineno-start: 11
-   :lines: 11-20
+   :lineno-start: 12
+   :lines: 12-23
    :linenos:
 
 This function simply reads a point cloud and calls ``draw_geometries_with_editing``. This function provides vertex selection and cropping.
@@ -65,8 +65,8 @@ The following script register two point clouds using point-to-point ICP. It gets
 
 .. literalinclude:: ../../../examples/Python/Advanced/interactive_visualization.py
    :language: python
-   :lineno-start: 43
-   :lines: 43-52
+   :lineno-start: 51
+   :lines: 51-60
    :linenos:
 
 The script reads two point clouds, and visualize the point clouds before alignment.
@@ -76,8 +76,8 @@ The script reads two point clouds, and visualize the point clouds before alignme
 
 .. literalinclude:: ../../../examples/Python/Advanced/interactive_visualization.py
    :language: python
-   :lineno-start: 30
-   :lines: 30-41
+   :lineno-start: 35
+   :lines: 35-48
    :linenos:
 
 Function ``pick_points(pcd)`` makes an instance of ``VisualizerWithEditing``. To mimic ``draw_geometries``, it creates windows, adds geometry, visualize geometry, and terminates. A novel interface function from ``VisualizerWithEditing`` is ``get_picked_points()`` that returns the indices of user-picked vertices.
@@ -115,8 +115,8 @@ Registration using user correspondences
 
 .. literalinclude:: ../../../examples/Python/Advanced/interactive_visualization.py
    :language: python
-   :lineno-start: 53
-   :lines: 53-71
+   :lineno-start: 61
+   :lines: 61-80
    :linenos:
 
 The later part of the demo computes an initial transformation based on the user-provided correspondences. This script builds pairs of correspondences using ``Vector2iVector(corr)``. It utilizes ``TransformationEstimationPointToPoint.compute_transformation`` to compute the initial transformation from the correspondences. The initial transformation is refined using ``registration_icp``.

docs/tutorial/Advanced/multiway_registration.rst

@@ -32,7 +32,7 @@ The first part of the tutorial script reads three point clouds from files. The p
 .. literalinclude:: ../../../examples/Python/Advanced/multiway_registration.py
    :language: python
    :lineno-start: 24
-   :lines: 24-58
+   :lines: 24-68
    :linenos:
 
 A pose graph has two key elements: nodes and edges. A node is a piece of geometry :math:`\mathbf{P}_{i}` associated with a pose matrix :math:`\mathbf{T}_{i}` which transforms :math:`\mathbf{P}_{i}` into the global space. The set :math:`\{\mathbf{T}_{i}\}` are the unknown variables to be optimized. ``PoseGraph.nodes`` is a list of ``PoseGraphNode``. We set the global space to be the space of :math:`\mathbf{P}_{0}`. Thus :math:`\mathbf{T}_{0}` is identity matrix. The other pose matrices are initialized by accumulating transformation between neighboring nodes. The neighboring nodes usually have large overlap and can be registered with :ref:`point_to_plane_icp`.
@@ -50,8 +50,8 @@ The script creates a pose graph with three nodes and three edges. Among the edge
 
 .. literalinclude:: ../../../examples/Python/Advanced/multiway_registration.py
    :language: python
-   :lineno-start: 72
-   :lines: 72-79
+   :lineno-start: 82
+   :lines: 82-89
    :linenos:
 
 Open3D uses function ``global_optimization`` to perform pose graph optimization. Two types of optimization methods can be chosen: ``GlobalOptimizationGaussNewton`` or ``GlobalOptimizationLevenbergMarquardt``. The latter is recommended since it has better convergence property. Class ``GlobalOptimizationConvergenceCriteria`` can be used to set the maximum number of iterations and various optimization parameters.
@@ -87,8 +87,8 @@ Visualize optimization
 
 .. literalinclude:: ../../../examples/Python/Advanced/multiway_registration.py
    :language: python
-   :lineno-start: 81
-   :lines: 81-85
+   :lineno-start: 91
+   :lines: 91-95
    :linenos:
 
 Ouputs:
@@ -105,8 +105,8 @@ Make a combined point cloud
 
 .. literalinclude:: ../../../examples/Python/Advanced/multiway_registration.py
    :language: python
-   :lineno-start: 87
-   :lines: 87-95
+   :lineno-start: 97
+   :lines: 97-106
    :linenos:
 
 .. image:: ../../_static/Advanced/multiway_registration/combined.png

docs/tutorial/Advanced/non_blocking_visualization.rst

@@ -51,7 +51,7 @@ Prepare example data
 .. literalinclude:: ../../../examples/Python/Advanced/non_blocking_visualization.py
    :language: python
    :lineno-start: 13
-   :lines: 13-28
+   :lines: 13-23
    :linenos:
 
 This part reads two point clouds and downsamples them. The source point cloud is intentionally transformed for the misalignment. Both point clouds are flipped for better visualization.
@@ -62,8 +62,8 @@ Initialize Visualizer class
 
 .. literalinclude:: ../../../examples/Python/Advanced/non_blocking_visualization.py
    :language: python
-   :lineno-start: 30
-   :lines: 30-33
+   :lineno-start: 25
+   :lines: 25-28
    :linenos:
 
 These lines make an instance of the visualizer class, open a visualizer window, and add two geometries to the visualizer.
@@ -73,8 +73,8 @@ Transform geometry and visualize it
 
 .. literalinclude:: ../../../examples/Python/Advanced/non_blocking_visualization.py
    :language: python
-   :lineno-start: 38
-   :lines: 38-48
+   :lineno-start: 33
+   :lines: 33-44
    :linenos:
 
 This script calls ``registration_icp`` for every iteration. Note that it explicitly forces only one ICP iteration via ``ICPConvergenceCriteria(max_iteration = 1)``. This is a trick to retrieve a slight pose update from a single ICP iteration. After ICP, source geometry is transformed accordingly.

docs/tutorial/Advanced/pointcloud_outlier_removal.rst

@@ -9,7 +9,7 @@ and artifact that one would like to remove. This tutorial address outlier remova
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
    :lineno-start: 5
-   :lines: 5-41
+   :lines: 5-
    :linenos:
 
 
@@ -20,8 +20,8 @@ A point cloud is loaded and downsampled using ``voxel_downsample``.
 
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
-   :lineno-start: 21
-   :lines: 21-27
+   :lineno-start: 22
+   :lines: 22-28
    :linenos:
 
 .. image:: ../../_static/Advanced/pointcloud_outlier_removal/voxel_down_sample.png
@@ -31,8 +31,8 @@ For comparison, ``uniform_down_sample`` can downsample point cloud by collecting
 
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
-   :lineno-start: 29
-   :lines: 29-31
+   :lineno-start: 30
+   :lines: 30-32
    :linenos:
 
 .. image:: ../../_static/Advanced/pointcloud_outlier_removal/uniform_down_sample.png
@@ -46,8 +46,8 @@ The selected points and the non-selected points are visualized.
 
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
-   :lineno-start: 9
-   :lines: 9-16
+   :lineno-start: 10
+   :lines: 10-17
    :linenos:
 
 
@@ -56,8 +56,8 @@ Statistical outlier removal
 
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
-   :lineno-start: 33
-   :lines: 33-36
+   :lineno-start: 34
+   :lines: 34-38
    :linenos:
 
 ``statistical_outlier_removal`` removes points that are further away from their neighbors compared to the average for the point cloud. It takes two input parameters:
@@ -73,8 +73,8 @@ Radius outlier removal
 
 .. literalinclude:: ../../../examples/Python/Advanced/pointcloud_outlier_removal.py
    :language: python
-   :lineno-start: 38
-   :lines: 38-41
+   :lineno-start: 40
+   :lines: 40-44
    :linenos:
 
 ``radius_outlier_removal`` removes points that have few neighbors in a given sphere around them. Two parameters can be used to tune the filter to your data: