update docs

Author: AtsushiSakai

ArmNavigation/two_joint_arm_to_point_control/Planar_Two_Link_IK.ipynb

@@ -4,7 +4,22 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Inverse Kinematics for a Planar Two-Link Robotic Arm\n",
+    "## Two joint arm to point control\n",
+    "\n",
+    "![TwoJointArmToPointControl](https://github.com/AtsushiSakai/PythonRobotics/raw/master/ArmNavigation/two_joint_arm_to_point_control/animation.gif)\n",
+    "\n",
+    "\n",
+    "\n",
+    "\n",
+    "This is two joint arm to a point control simulation.\n",
+    "\n",
+    "This is a interactive simulation.\n",
+    "\n",
+    "You can set the goal position of the end effector with left-click on the ploting area.\n",
+    "\n",
+    "\n",
+    "\n",
+    "### Inverse Kinematics for a Planar Two-Link Robotic Arm\n",
     "\n",
     "A classic problem with robotic arms is getting the end-effector, the mechanism at the end of the arm responsible for manipulating the environment, to where you need it to be. Maybe the end-effector is a gripper and maybe you want to pick up an object and maybe you know where that object is relative to the robot - but you cannot tell the end-effector where to go directly. Instead, you have to determine the joint angles that get the end-effector to where you want it to be. This problem is known as inverse kinematics.\n",
     "\n",
@@ -403,7 +418,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.5.2"
+   "version": "3.6.6"
   }
  },
  "nbformat": 4,

Localization/Kalmanfilter_basics.ipynb

@@ -4,23 +4,23 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# _KF Basics - Part I_\n",
-    "----\n"
+    "## KF Basics - Part I\n"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "##### What is the need to describe belief in terms of PDF's?\n",
+    "### Introduction\n",
+    "#### What is the need to describe belief in terms of PDF's?\n",
     "This is because robot environments are stochastic. A robot environment may have cows with Tesla by side. That is a robot and it's environment cannot be deterministically modelled(e.g as a function of something like time t). In the real world sensors are also error prone, and hence there'll be a set of values with a mean and variance that it can take. Hence, we always have to model around some mean and variances associated."
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "##### What is Expectation of a Random Variables?\n",
+    "#### What is Expectation of a Random Variables?\n",
     " Expectation is nothing but an average of the probabilites\n",
     " \n",
     "$$\\mathbb E[X] = \\sum_{i=1}^n p_ix_i$$\n",
@@ -56,7 +56,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "##### What is the advantage of representing the belief as a unimodal as opposed to multimodal?\n",
+    "#### What is the advantage of representing the belief as a unimodal as opposed to multimodal?\n",
     "Obviously, it makes sense because we can't multiple probabilities to a car moving for two locations. This would be too confusing and the information will not be useful."
    ]
   },
@@ -71,9 +71,9 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Variance, Covariance and Correlation\n",
-    "----\n",
-    "### Variance\n",
+    "### Variance, Covariance and Correlation\n",
+    "\n",
+    "#### Variance\n",
     "Variance is the spread of the data. The mean does'nt tell much **about** the data. Therefore the variance tells us about the **story** about the data meaning the spread of the data.\n",
     "\n",
     "$$\\mathit{VAR}(X) = \\frac{1}{n}\\sum_{i=1}^n (x_i - \\mu)^2$$"
@@ -104,7 +104,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### Covariance\n",
+    "#### Covariance\n",
     "\n",
     "This is for a multivariate distribution. For example, a robot in 2-D space can take values in both x and y. To describe them, a normal distribution with mean in both x and y is needed.\n",
     "\n",
@@ -224,10 +224,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Gaussians \n",
-    "----\n",
-    "\n",
-    "\n",
+    "### Gaussians \n",
     "\n",
     "#### Central Limit Theorem\n",
     "\n",
@@ -285,7 +282,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "##### Gaussian Distribution\n",
+    "#### Gaussian Distribution\n",
     "A Gaussian is a *continuous probability distribution* that is completely described with two parameters, the mean ($\\mu$) and the variance ($\\sigma^2$). It is defined as:\n",
     "\n",
     "$$ \n",
@@ -337,7 +334,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "##### Why do we need Gaussian distributions?\n",
+    "#### Why do we need Gaussian distributions?\n",
     "Since it becomes really difficult in the real world to deal with multimodal distribution as we cannot put the belief in two seperate location of the robots. This becomes really confusing and in practice impossible to comprehend. \n",
     "Gaussian probability distribution allows us to drive the robots using only one mode with peak at the mean with some variance."
    ]
@@ -346,8 +343,8 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## Gaussian Properties\n",
-    "----\n",
+    "### Gaussian Properties\n",
+    "\n",
     "**Multiplication**\n",
     "\n",
     "\n",
@@ -569,7 +566,7 @@
     "\n",
     "\n",
     "\n",
-    "##### numpy einsum examples"
+    "** numpy einsum examples **"
    ]
   },
   {
@@ -757,7 +754,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.5"
+   "version": "3.6.6"
   }
  },
  "nbformat": 4,

Localization/Kalmanfilter_basics_2.ipynb

@@ -4,7 +4,19 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Extended Kalman Filter Localization\n",
+    "## Extended Kalman Filter Localization\n",
+    "\n",
+    "![EKF](https://github.com/AtsushiSakai/PythonRobotics/raw/master/Localization/extended_kalman_filter/animation.gif)\n",
+    "\n",
+    "This is a sensor fusion localization with Extended Kalman Filter(EKF).\n",
+    "\n",
+    "The blue line is true trajectory, the black line is dead reckoning\n",
+    "trajectory,\n",
+    "\n",
+    "the green point is positioning observation (ex. GPS), and the red line\n",
+    "is estimated trajectory with EKF.\n",
+    "\n",
+    "The red ellipse is estimated covariance ellipse with EKF.\n",
     "\n",
     "Code; [PythonRobotics/extended\\_kalman\\_filter\\.py at master · AtsushiSakai/PythonRobotics](https://github.com/AtsushiSakai/PythonRobotics/blob/master/Localization/extended_kalman_filter/extended_kalman_filter.py)"
    ]
@@ -13,15 +25,24 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Kalman Filter"
+    "### Kalman Filter"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Ref"
+    "### Ref\n",
+    "\n",
+    "- [PROBABILISTIC\\-ROBOTICS\\.ORG](http://www.probabilistic-robotics.org/)"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {

…on/extended_kalman_filter/notebook.ipynb …xtended_kalman_filter_localization.ipynbLocalization/extended_kalman_filter/notebook.ipynb renamed to Localization/extended_kalman_filter/extended_kalman_filter_localization.ipynb Localization/extended_kalman_filter/notebook.ipynb renamed to Localization/extended_kalman_filter/extended_kalman_filter_localization.ipynb

@@ -40,6 +40,7 @@ See this paper for more details:
    modules/path_tracking
    modules/arm_navigation
    modules/aerial_navigation
+   modules/appendix
 
 
 Indices and tables

docs/index.rst

@@ -23,19 +23,35 @@ def generate_rst(npath):
     # print(npath)
 
     # generate dir
-    dirpath = "./modules/" + os.path.dirname(npath)[3:]
+    dirpath = os.path.dirname(npath)
     # print(dirpath)
 
-    if not os.path.exists(dirpath):
-        os.makedirs(dirpath)
-
     rstpath = os.path.abspath("./modules/" + npath[3:-5] + "rst")
-    print(rstpath)
+    # print(rstpath)
+
+    basename = os.path.basename(rstpath)
 
     cmd = "jupyter nbconvert --to rst "
     cmd += npath
-    cmd += " --output "
-    cmd += rstpath
+    print(cmd)
+    subprocess.call(cmd, shell=True)
+
+    cmd = "rm -rf "
+    cmd += "./modules/"
+    cmd += basename[:-4]
+    cmd += "*"
+    print(cmd)
+    subprocess.call(cmd, shell=True)
+
+    cmd = "mv "
+    cmd += dirpath
+    cmd += "/*.rst ./modules/"
+    print(cmd)
+    subprocess.call(cmd, shell=True)
+
+    cmd = "mv "
+    cmd += dirpath
+    cmd += "/*_files ./modules/"
     print(cmd)
     subprocess.call(cmd, shell=True)