Category: Python

[Trouble Shooting] NullFunctionError of glutInit in OpenGL


When trying to call “glutInit()” of PyOpenGL, I got the error “NullFunctionError: Attempt to call an undefined function glutInit, check for bool(glutInit) before calling”.

# python source code
from OpenGL.GL import *
from OpenGL.GLUT import *
from OpenGL.GLU import *

glutInitDisplayMode(GLUT_SINGLE | GLUT_RGB) 
Error Detail

NullFunctionError                         Traceback (most recent call last)
/tmp/ipykernel_14/ in <module>
      3 from OpenGL.GLU import *
----> 5 glutInit()
      6 glutInitDisplayMode(GLUT_SINGLE | GLUT_RGB)

/usr/local/lib/python3.9/site-packages/OpenGL/GLUT/ in glutInit(*args)
    331     try:
    332         # XXX need to check for error condition here...
--> 333         _base_glutInit( ctypes.byref(count), holder )
    334     finally:
    335         os.chdir( currentDirectory )

/usr/local/lib/python3.9/site-packages/OpenGL/platform/ in __call__(self, *args, **named)
    421                 pass
    422             else:
--> 423                 raise error.NullFunctionError(
    424                     """Attempt to call an undefined function %s, check for bool(%s) before calling"""%(
    425                         self.__name__, self.__name__,

NullFunctionError: Attempt to call an undefined function glutInit, check for bool(glutInit) before calling


Python 3.9.7


There are some Recommended Enhancements for PyOpenGL. And some of them, such as GLUT or FreeGLUT, are not contained in PyOpenGL itself.


Solution1 use binary installer for windows.

The documentation says “Win32 and Win64 binary installers for PyOpenGL include a copy of GLUT”.
So uninstall current pyopengl first.
And download binary installer for windows. Put the downloaded file in some directory, then run the command “pip install <file name>” in the directory where the file is saved.

pip uninstall PyOpenGL
pip install PyOpenGL‑3.1.6‑cp39‑cp39‑win_amd64.whl

Solution2 install GLUT or FreeGLUT separately

Installing and adding GLUT or FreeGLUT to PATH solves this problem.

Build FreeGLUT from source code or download built FreeGLUT from “Download freeglut 3.0.0 for MSVC” in

Put freeglut\bin\x64\freeglut.dll into C:\Windows\System32.

Add “C:\Windows\System32” to environment variable “PATH” to enable python find the library.

If the error still occurred, check if your library name is in Win32Platform.GLUT() <python>\Lib\site-packages\OpenGL\platform\

    def GLUT( self ):
        for possible in ('freeglut%s.%s'%(size,vc,), 'freeglut', 'glut%s.%s'%(size,vc,)):  # Added 'freeglut' because the library name is freeglut.dll
            # Prefer FreeGLUT if the user has installed it, fallback to the included 
            # GLUT if it is installed
                return ctypesloader.loadLibrary(
                    ctypes.windll, possible, mode = ctypes.RTLD_GLOBAL
            except WindowsError:
        return None

These links might help you.
python – Attempt to call an undefined function glutInit
#219 glutInit fails on windows 64-bit
Attempt to call an undefined function glutInit
Python and PyOpenGL Installation

Appendix: How to install GLUT or FreeGLUT on Linux

Use “sudo apt-get install python-opengl” command to install PyOpenGL with dependent library at once.

How To Get Mapping of UV and Vertex


To get correspondence of UV coordinates and vertex id in the object in Blender scene by Blender Python. The following is output of the script.

uv: 0.625 0.5 vertex id: 0
uv: 0.875 0.5 vertex id: 4
uv: 0.875 0.75 vertex id: 6
uv: 0.625 0.75 vertex id: 2
uv: 0.375 0.75 vertex id: 3


Blender 2.83 (Python 3.7.4)
Windows 10


Mesh Loop

Use Mesh Loop to get UV and vertex data.

In unfolded UV, the correspondence between the point of UV map and the vertex of 3D object is not one-to-one. Thus we use ‘Mesh Loop’ instead of vertex or edge to distinct point of UV map.

Mesh Loop is an element surrounding a face, which consists of edges and vertices. Each quad face consists of 4 Mesh Loops as below. For example, cube objects have 6face, 24 mesh loops.

import bpy

obj =[obj_name]
mesh_loops = 
for mesh_loop in mesh_loops:

Mesh UV Loop

UVs have Mesh UV Loops as a counterpart to the Mesh Loops. And the correspondence between Mesh Loop and MeshUVLoop is one-to-one.

import bpy

obj =[obj_name]
mesh_uv_loop =[0].data[loop_index]

Source Code

import bpy

def get_uvs(obj_name):
    obj =[obj_name]
    mesh_loops =

    for i, mesh_loop in enumerate(mesh_loops):
        mesh_uv_loop =[0].data[i]
        print("uv:", mesh_uv_loop.uv[0], mesh_uv_loop.uv[1], "vertex id:", mesh_loop.vertex_index)


How To Start Anaconda


Today’s goal is to construct development environment for data analytics and machine learning with Anaconda.

What is Anaconda?

Anaconda is open source Python distribution for data science. You can see the list of package lists here.

the open-source Individual Edition (Distribution) is the easiest way to perform Python/R data science and machine learning on a single machine. Developed for solo practitioners, it is the toolkit that equips you to work with thousands of open-source packages and libraries.



macOS Catalina 10.15.5
Python 3.9.7


Download & Install

Access the website and click “Download” button to download the installer of the edition that you want.

Start the installer and click “Continue”.

Check if anaconda is completely installed with terminal.

Start Anaconda Navigator

Start applications > Anaconda-Navigator.

Create virtual environment

Click Environments and create new environment.

I named new environment “data_analysis”.

Install libraries or modules with conda

Conda is an open source package management system and environment management system.

Open terminal in the environment where you want to install libraries.

Then put the “conda install” command to install libraries.
For example, I installed pytorch in the “pytorch” environment. The option “-c” is the channel (What is a “conda channel”?).

conda install pytorch torchvision -c pytorch

Start Application

Select environment what you want to use and install or launch application.

I launched “Jupyter notebook” and check that “pythorch” library is installed successfully.

What Is Convolution In Neural Network?


Today’s goal is to understand a layer of process “convolution” in neural network. In other words, this article describes convolutional neural network (CNN or ConvNet) that is a neural network contains convolutional layers in its architecture.

What Convolution Is

Convolution is the binary operation to takes 2 functions and produces a function \(f * g\), which is defined as the integral of the product of the two functions after one is reversed and shifted.

$$(f*g)(t) = \int f(\tau)g(t – \tau)d\tau $$

Examples of 2 function

In the context of image processing, convolution is the process of generating output image by weighting input image with the filter. For example, convolution is used in smoothing with gaussian filter.

Example of convolution by Gaussian filter

The image of weights is called “kernel” in convolution. And the output is the compression of input.

In CNN, the convolution layer is used to generate output image that represents feature of input image while compressing local region. 1 layer can have several kernels and the output is the same number of feature maps as kernels.

Architecture of CNN
by Aphex34

Details Of Convolution Layer

In convolution layer, the input image will be converted. For example, a 5*5 kernel is applied to RGB 32*32 image to generate 28*28 output as below.

Note that there are 2 styles of convolution for RGB image. The first one is the way in which 3 maps generated from RGB channels by convolution are added together to 1 channel output image. In the second way, each channel is converted to each channel of output image. In the latter case the output image has 3 channels.

The former is commonly used in CNN.

If the convolution layer has 5 different kernels, the output has 5 channels.


Parameters that determines convolution process.

  • channel
  • kernel size
  • stride
  • padding
  • dilation

Channel and kernel size

channel = input channel * the number of kernel. The channel of output is the same as the channel of kernel.

Kernel size is the size of kernel side.


Stride is the size that filter sliding to input image.


Padding is the amount of padding space applied on both sides of input. Padding changes output size to fit input image size.


Dilation controls the spacing between the kernel points.

Convolution In PyTorch

PytTorch provides a method torch.nn.Conv2d() for 2D convolution. Conv2d makes a convolutional layer in network class that extends nn.Module as below.

import torch
import torch.nn as nn
import torch.nn.functional as F

class Net(nn.Module):
  def __init__(self):
    super(Net, self).__init__()
    self.conv1 = nn.Conv2d(1, 6, 3) #1channel input, 6channel output (6kernels), 3*3 kernel
    self.conv2 = nn.Conv2d(6, 16, 3) #6channel input, 16channel output, 3*3 kernel 

    self.fc1 = nn.Linear(16 * 6 * 6, 120) #Linear transform from 16*6*6 features into 84 features
    self.fc2 = nn.Linear(120, 84)
    self.fc3 = nn.Linear(84, 10)

  def forward(self, x):
    x = F.relu(self.conv1(x))
    x = F.max_pool2d(x, (2, 2))
    x = F.relu(self.conv2(x))
    x = F.max_pool2d(x, 2)
    x = x.view(-1, self.num_flat_features(x))
    x = F.relu(self.fc1(x))
    x = F.relu(self.fc2(x))
    x = self.fc3(x)
    return x

  def num_flat_features(self, x):
    size = x.size()[1:]
    num_features = 1
    for s in size:
      num_features *= s
    return num_flat_features

net = Net()

See the detail of net architecture by print().


  (conv1): Conv2d(1, 6, kernel_size=(3, 3), stride=(1, 1))
  (conv2): Conv2d(6, 16, kernel_size=(3, 3), stride=(1, 1))
  (fc1): Linear(in_features=576, out_features=120, bias=True)
  (fc2): Linear(in_features=120, out_features=84, bias=True)
  (fc3): Linear(in_features=84, out_features=10, bias=True)

See the size of parameters in conv1 and conv2.

params = list(net.parameters())

torch.Size([6, 1, 3, 3])
torch.Size([16, 6, 3, 3])

[Trouble Shooting] Cannot Use Library In Jupyter Notebook


Python 3.8.4
pip 20.1.1


I installed “matblotlib” by using pip3 command on the terminal as below. I can import matplotlib on the terminal.

% pip3 install matplotlib 
% python3
Python 3.8.4 (default, Jul 14 2020, 02:58:48) 
[Clang 11.0.3 (clang-1103.0.32.62)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import matplotlib

In jupyter notebook,

import matplotlib
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-3-0484cd13f94d> in <module>
----> 1 import matplotlib

ModuleNotFoundError: No module named 'matplotlib'

Check the situation

The reference path for jupyter notebook might differ from the one of host python3.

Jupyter notebook refers the python 3.8 in jupyterlab/2.2.0/libexec/lib

import sys

3.8.4 (default, Jul 14 2020, 02:58:48) 
[Clang 11.0.3 (clang-1103.0.32.62)]
['/usr/local/Cellar/jupyterlab/2.2.0/libexec/lib/', '/usr/local/Cellar/jupyterlab/2.2.0/libexec/lib/python3.8', '/usr/local/Cellar/jupyterlab/2.2.0/libexec/lib/python3.8/lib-dynload', '/usr/local/opt/python@3.8/Frameworks/Python.framework/Versions/3.8/lib/python3.8', '', '/usr/local/Cellar/jupyterlab/2.2.0/libexec/lib/python3.8/site-packages', '/usr/local/Cellar/jupyterlab/2.2.0/libexec/lib/python3.8/site-packages/IPython/extensions', '/Users/<username>/.ipython']

On terminal, Python3 is in /usr/local/Cellar.

>>> import sys
>>> print(sys.version)
>>> print(sys.path)

3.8.4 (default, Jul 14 2020, 02:58:48) 
[Clang 11.0.3 (clang-1103.0.32.62)]

['', '/usr/local/Cellar/python@3.8/3.8.4/Frameworks/Python.framework/Versions/3.8/lib/', '/usr/local/Cellar/python@3.8/3.8.4/Frameworks/Python.framework/Versions/3.8/lib/python3.8', '/usr/local/Cellar/python@3.8/3.8.4/Frameworks/Python.framework/Versions/3.8/lib/python3.8/lib-dynload', '/usr/local/lib/python3.8/site-packages']


In my case, the problem caused by homebrew. Installation of Jupyter Lab or Jupyter Notebook with brew command caused the different reference to python3 as above.

We can check the list of installed items with the following command on the terminal.

% brew list
autoconf	ipython		openssl@1.1	python@3.8	xz
flake8		jupyterlab	pandoc		readline	zeromq
gdbm		libyaml		perl		ruby
gettext		lua		pkg-config	sqlite
icu4c		node		pyenv		vim


My solutions is to uninstall jupyterlab with brew command and install it again with pip3.

% brew uninstall jupyterlab
Uninstalling /usr/local/Cellar/jupyterlab/2.2.0... (7,150 files, 106.8MB)

% python3 -m pip install jupyterlab
Collecting jupyterlab
  Downloading jupyterlab-3.1.7-py3-none-any.whl (8.5 MB)
     |████████████████████████████████| 8.5 MB 136 kB/s 

Start Jupyter Lab from the terminal.

% jupyter lab

I found another it could be a solution for similar problem to install ipykernel. Check ModuleNotFound Error for matplotlib for detail.

Blender Addon Naming Rules


Today’s goal is to list naming rules and conventions of Blender addon.

AddonPreferences(bpy_struct) from Blender 2.93.2 Release Candidate Python API documentation
Script Meta Info from


Blender 2.83 (Python 3.7.4)

directory_name and

Addon directory and files are snake_case with only lower cases. See examples in your addon directory, C:/Program Files/Blender Foundation/Blender 2.83/2.83/scripts/addons.

Addon Name

Addon name defined in bl_info is defined with the first letter of each word capitalized.

Class Name

2.8x enforces naming conventions for class name. Refer to “Class Registration” > “Naming” in “Blender 2.80: Addon API” reference for details.

The naming convention is UPPER_CASE_{SEPARATOR}_mixed_case. And {SEPARATOR} is defined according to the type of class as below.

  • Header -> _HT_
  • Menu -> _MT_
  • Operator -> _OT_
  • Panel -> _PT_
  • UIList -> _UL_

The followings are examples of panel class name.

# Header
# from "oscurart_tools" addon
class OSSELECTION_HT_OscSelection(bpy.types.Header):
    bl_label = "Selection Osc"
    bl_space_type = "VIEW_3D"

# Menu
# from "space_view3d_spacebar_menu" addon
class VIEW3D_MT_Space_Dynamic_Menu(Menu):
    bl_label = "Dynamic Context Menu"

class VIEW3D_MT_View_Menu(Menu):
    bl_label = "View"

# Operator
# from "add_camera_rigs" addon
class ADD_CAMERA_RIGS_OT_set_scene_camera(Operator):
    bl_idname = "add_camera_rigs.set_scene_camera"
    bl_label = "Make Camera Active"
    bl_description = "Makes the camera parented to this rig the active scene camera"

# Panels
# from "add_camera_rigs" addon
class ADD_CAMERA_RIGS_PT_camera_rig_ui(Panel, CameraRigMixin)
    bl_label = "Camera Rig"
    bl_space_type = 'VIEW_3D'
    bl_region_type = 'UI'
    bl_category = 'Item'

# UIList
# from "cycles" addon
class CYCLES_RENDER_UL_aov(bpy.types.UIList):
    def draw_item(self, ...


bl_idname is specified to access the operator from python script in Blender. The bl_idname consists of snake_case words connected with dots.

# bl_idname in cycle/ in cycle addon

class CYCLES_OT_use_shading_nodes(Operator):
    bl_idname = "cycles.use_shading_nodes"

class CYCLES_OT_add_aov(bpy.types.Operator):

class CYCLES_OT_remove_aov(bpy.types.Operator):

class CYCLES_OT_denoise_animation(Operator):
    bl_idname = "cycles.denoise_animation"

class CYCLES_OT_merge_images(Operator):
    bl_idname = "cycles.merge_images"

 For headers, menus and panels, the bl_idname is expected to match the class name (automatic if none is specified). Refer to “Class Registration” > “Naming” in “Blender 2.80: Addon API” reference.

Other – PEP 8

Basically the naming rules of module, variable and so on is compliant with PEP 8.

variable = "hello"
variable_two = 2

def function_name(arg_name):

class ClassName():
    def __init__(self, arg_name):
        self.variable = 1
    def method_name(self):

How To Make My Addon Support “Undo” In Blender


Today’s goal is to implement “Undo” in my Blender addon.


Blender 2.83


Add bl_options “UNDO” or “UNDO_GROUPED” into the your custom operator. It’s easy.

UNDO Undo: Push an undo event (needed for operator redo).
UNDO_GROUPED: Grouped Undo, Push a single undo event for repeated instances of this operator.

from Blender2.93.2 Python API Documentation

The following is an example.

class ADDMATRIX_add_cube(bpy.types.Operator):
    bl_idname = 'add_matrix_obj.add_cube'
    bl_label = "Add matrix cube"
    bl_options = {'REGISTER', "UNDO"} # Add "UNDO" option here

    input1: bpy.props.IntProperty()
    input2: bpy.props.IntProperty()

    def execute(self, context):
        for xi in range(self.input1):
            x = xi*1.2
            for yi in range(self.input2):
                y = yi*1.2
                bpy.ops.mesh.primitive_cube_add(size=0.5, enter_editmode=False, align='WORLD', location=(x, y, 0))
        return {'FINISHED'}


be aware of the error where Blender crashed when undo after executing script in python console. (Reference: Fix T86293: crash undoing after executing the python console in certain)

That was fixed in Blender 2.93, the latest version. And You can fix the error by adding bl_options “UNDO_GROUPED” into the ConsoleExec operation in C:/Program Files/Blender Foundation/Blender 2.83/2.83/scripts/startup/bl_operators/

class ConsoleExec(Operator):
    """Execute the current console line as a python expression"""
    bl_idname = "console.execute"
    bl_label = "Console Execute"
    bl_options = {"UNDO_GROUPED"} # add undo here

    interactive: BoolProperty(

    def poll(cls, context):
        return (context.area and context.area.type == 'CONSOLE')

How To Get Icon List In Blender


Today’s goal is to show Icon Viewer that shows the list of icon we can use in Blender and use it.

Icon Viewer




1. Activate addon “Development Icon Viewer”

2. Open Icon Viewer

Open text Editor and click “Dev” tab.

Click on any icon you like and it’s name will be copied to the clipboard. Then paste it into your editor.


1. Operator button

def draw(self, context):
    layout = self.layout
    col = layout.column(align=True)
    col.operator("mesh.primitive_monkey_add", icon="MONKEY")

2. EnumProperty

class MyProp(bpy.types.PropertyGroup):
    items = [("id1", "name1", "description 1", "HELP", 0),
             ("id2", "name2", "description 2", "GHOST_ENABLED", 1),
             ("id3", "mame3", "description 3", "FUND", 2),]
    enum: bpy.props.EnumProperty(items=items)
def draw(self, context):
    layout = self.layout
    col = layout.column(align=True)
    col.prop(context.scene.MyProp, "enum")

How To Pass Arguments To Custom Operator In Blender Python


Today’s goal is to create operator that takes arguments in Blender Python.

The following is a custom operator that takes 2 argument “count_x” and “count_y”, and add cube object in the form of count_x rows and count_y columns.


Blender 2.83(LTS)


1. Create operator with properties

Pass the argument values through the properties to custom operator. Use “:” to add property in Blender custom operator as below. Check “Operator Example” in Blender manual for details.

class ADDMATRIX_add_cube(bpy.types.Operator):
    bl_idname = 'add_matrix_obj.add_cube'
    bl_label = "Add matrix cube"
    bl_options = {'REGISTER', "UNDO"}

    input1: bpy.props.IntProperty() # add argument1 as property "input1"
    input2: bpy.props.IntProperty() # add argument2 as property "input2"

    def execute(self, context):
        for xi in range(self.input1):
            x = xi*1.2
            for yi in range(self.input2):
                y = yi*1.2
                bpy.ops.mesh.primitive_cube_add(size=0.5, enter_editmode=False, align='WORLD', location=(x, y, 0))
        return {'FINISHED'}

You can execute operator with passing values of arguments “input1” and “input2” as below.


List Of Blender Properties


Today’s goal is to list up property UI that can be used in Blender Addon.
Reference: Property Definitions


Blender 2.83(LTS)

List of Properties

  • BoolProperty
  • BoolVectorProperty
  • CollectionProperty
  • EnumProperty
  • FloatProperty
  • FloatVectorProperty
  • IntProperty
  • IntVectorProperty
  • PointerProperty
  • StringProperty

How each property looks like


bool: bpy.props.BoolProperty()


bool_vector: bpy.props.BoolVectorProperty()

subtypes of BoolVectorProperty

The UI style or label differs according to its subtype.

bool_vector_color: bpy.props.BoolVectorProperty(subtype='COLOR')
bool_vector_translation: bpy.props.BoolVectorProperty(subtype='TRANSLATION')
# and more...


UILayout.prop() just shows the number of CollectionProperty. (Use “template_list” to show the item list of CollectionProperty)

class PropertySetTest(bpy.types.PropertyGroup):
    test_int: bpy.props.IntProperty()
    test_string: bpy.props.StringProperty()
collection: bpy.props.CollectionProperty(type=PropertySetTest)


items = [("id1", "name1", "description 1", "MESH_CUBE", 0),
         ("id2", "name2", "description 2", "MESH_TORUS", 1),
         ("id3", "mame3", "description 3", "CONE", 2),]
enum: bpy.props.EnumProperty(items=items)


float: bpy.props.FloatProperty()

subtypes of FloatProperty

float_pixel: bpy.props.FloatProperty(subtype="PIXEL")
float_unsigned: bpy.props.FloatProperty(subtype="UNSIGNED")
# and more...

units of FloatProperty

float_none: bpy.props.FloatProperty(unit='NONE')
float_length: bpy.props.FloatProperty(unit='LENGTH')
# and more...


float_vector: bpy.props.FloatVectorProperty()

subtypes of FloatVectorProperty

units of FloatVectorProperty


subtypes of IntVectorProperty


subtypes of FloatVectorProperty


PointerProperty takes argument ‘type’ to set the type of pointer.

pointer_scene: bpy.props.PointerProperty(type=bpy.types.Scene)
pointer_object: bpy.props.PointerProperty(type=bpy.types.Object)
pointer_matarial: bpy.props.PointerProperty(type=bpy.types.Material)
pointer_mesh: bpy.props.PointerProperty(type=bpy.types.Mesh)
pointer_image: bpy.props.PointerProperty(type=bpy.types.Image)
pointer_brush: bpy.props.PointerProperty(type=bpy.types.Brush)
pointer_camera: bpy.props.PointerProperty(type=bpy.types.Camera)
pointer_light: bpy.props.PointerProperty(type=bpy.types.Light)


subtypes of FloatVectorProperty