Lecture17_04 - Exploring Hamiltonian Monte Carlo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploring Hamiltonian Monte Carlo\n",
    "\n",
    "##### Keywords: hamiltonian monte carlo, energy, NUTS, leapfrog, canonical distribution, microcanonical distribution, transition distribution, marginal energy distribution, data augmentation, classical mechanics, detailed balance, statistical mechanics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "pd.set_option('display.width', 500)\n",
    "pd.set_option('display.max_columns', 100)\n",
    "import seaborn as sns\n",
    "\n",
    "from IPython.core.display import Image\n",
    "\n",
    "import pymc3 as pm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction \n",
    "\n",
    "A lot of HMC's recent appreciation as an MCMC method is due to [Neal](https://arxiv.org/pdf/1206.1901.pdf), and we'll follow him and [Betancourt](https://arxiv.org/abs/1701.02434) heavily here. All the signature Maroon diagrams are from Betancourt's review paper."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Looking for the perfect MCMC?\n",
    "\n",
    "The problem with MH is that is sensitive to the step size, Gibbs we need to know how to draw from the conditionals. What we want is the ability to adjust the step size but at the same time preserving the properties of MCMC, namely being able to sample the whole space and at the same time converge to the target distribution. We also would like to have no burnin, no autocorrelation, guarantee mixing and convergence very fast and no tunable parameters (and a pony). \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In other words, we'd like to explore the typical-set surface smoothly. To do this we must first characterize the surface, something we can do via a gradient. To do this we must identify the equation of the typical set surface so that we can find the gradient which is perpendicular to the surface. And once we do that, we are not done, as the gradient points towards regions of higher density (modes) from the surface of the typical set.\n",
    "\n",
    "To do this imagine a sliver $dP = p(q)dq$ thats in the typical set. If $dq$ (we are using $q$ instead of $x$) is small enough, then we can consider the typical set as a collection of foliates $\\{q_i\\}$ each of constant probability density  $p(q)=c_i$ where $c_i$ is a constant. Thus there are n such foliates, or \"orbits\", or level sets. Now we know that the gradient is perpendicular to such level-sets and we can use it to characterize these sets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mechanics to the rescue\n",
    "\n",
    "We can make our usual connection to the energy of a physical system by tentatively identifying the energy $E = - log(p(q))$. This is what we did to find distributions in Metropolis and the inverse $p(q) = e^{-E(q)/T}$ the minima of functions in simulated annealing. There we proposed a move using a proposal distribution, creating a random walk.\n",
    "\n",
    "We dont want to do that here, preferring something that will move us smoothly along a level set. We use our newly acquired knowledge of data-augmentation and gibbs-sampling from an augmented distribution instead.\n",
    "\n",
    "But first let us make a change of purpose and notation. Lets write now the potential energy as:\n",
    "\n",
    "$$V(q) = - log\\, p(q)$$\n",
    "\n",
    "such that\n",
    "\n",
    "$$p(q) = e^{-V(q)}.$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1188db6a0>]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1186483c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fun = lambda q: sp.stats.norm.pdf(q) \n",
    "\n",
    "qq = np.linspace(-5,5,100)\n",
    "plt.plot(qq, fun(qq))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x117c36eb8>]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAECCAYAAADjBlzIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhBAWJ0UuhBAWJ0UuhBAWJ0UuhBAW9/8Bi7/kZkaaxqkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118a36438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "V = lambda q: -np.log(fun(q))\n",
    "plt.plot(qq, V(qq))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The basic idea is to add a **momentum** variable $p$ for each $q$ in our probability density, adding a kinetic energy term to the potential energy to create a total energy, and thus creating a joint pdf $p(p,q)$. \n",
    "\n",
    "How would this work? And why momentum? Lets think about a rocket (or a satellite with thrusters it can fire) in orbit around the earth\n",
    "\n",
    "![](../images/rocketorbit.png)\n",
    "\n",
    "If this rocket had no velocity, it would simply fall down to the earth because it would not be able to counterbalance earth's gravitational potential (the equivalent of the energy we formulated above and have used in simulated annealing). On the other hand, if it had too much velocity, it would escape earth's gravity and take off to mars or similar.\n",
    "\n",
    "If we add just the right amount of momentum, it will exactly counterbalance the gravitational force, and the satellite will continue to move in its orbit. And this is an orbit of minimum energy (and therefore constant energy, since a system at minumum energy wont lift from it unless kicked(perhaps stochastically) to do so). Thus the satellite will move exactly along a level-curve of the energy level, in a direction exactly perpendicular to the gradient. Such motion is called **conservative**, as it conserves energy. In mechanics, the conserving of energy is related to the time-independence and thus time irreversibility of motion: is there is no *explicit* dependence on time in the equations of motion (there always is implicit dependence as positions and momenta depend on time), energy is conserved."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Augmentation\n",
    "\n",
    "Recall that the basic idea behind Data Augmentation is to resolve difficulty in simulate from $p(x)$ using the traditional methods by constructing a joint pdf such that:\n",
    "\n",
    "$$ \\int p(x,y)\\, dy = p(x)  $$\n",
    "\n",
    "and simulating from the associated conditional pdfs, $p(x|y)$ and $p(y|x)$ is easy.\n",
    "\n",
    "The DA algorithm is based on this joint pdf, and we *marginalize* over the **augmenting variable** $y$ to obtain samples for $p(x)$.\n",
    "\n",
    "The simplest form of DA algorithm takes the following form:\n",
    "\n",
    "1. Draw $Y\\sim P_{Y|X}(.|x)$ and call the observed value y\n",
    "2. Draw $X_{n+1} \\sim P_{X|Y}(.|y)$\n",
    "\n",
    "After disregarding the $Y_i$ in our samples in making a $X$ histogram, we have  samples $X_i \\sim P(x)$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To achieve our goal we introduce a new variable as we did for DA and Slice Sampling we now call $p$. As we explained in DA the new joint distribution \n",
    "\n",
    "$$p(q) = \\int p(q,p) \\, dp $$\n",
    "\n",
    "An easy way to achieve this is if $p(q,p)=p(q)p(p \\vert q)$. And since we can choose our joint distribution as we like, we choose (the reasons will be obvious soon). \n",
    "\n",
    "$$ p(q,p) = e^{-V(q)-K(q,p)} = e^{-V(q)} e^{-p^2/2m}$$\n",
    "\n",
    "Where $m$ is just a constant thus  $p(p)$ is  Gaussian. The kinetic energy if $p$ is the momentum is $K(q, p)=p^2/2m$.  This kinetic energy $K(p,q)$ has a mass parameter $m(q)$(one can consider generalizations of the kinetic energy term from this version we have from introductory physics).\n",
    "\n",
    "The choice of a kinetic energy term then is the choice of a conditional probability distribution over the \"augmented\" momentum which ensures that\n",
    "\n",
    "$$\\int dp p(p, q) = \\int dp p(p \\vert q) p(q) = p(q) \\int p(p \\vert q) dp = p(q).$$\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Thus the key to moving a sampler along a probability level curve is to give the sampler momentum and thus kinetic energy via the augmented momemtum variable. In other words, we must carry out an augmentation with an additional momentum which leaves the energy **Hamiltonian**\n",
    "\n",
    "$$H(p, q) = \\frac{p^2}{2m} +  V(q) = E_i,$$\n",
    "\n",
    "and thus:\n",
    "\n",
    "$$H(p,q) = -log(p(p,q)) = -log p(p \\vert q) - log p(q)$$\n",
    "\n",
    "with $E_i$ constants (constant energies) for each level-set foliate and where the **potential energy** $V(q) = -log(p(q))$ replaces the energy term we had earlier in simulated annealing. \n",
    "\n",
    "**Thus it is critical to realize that we are now looking at level curve foliates of $p(p,q)$, not those of $p(q)$**. Our real aim is to explore the latter. This joint distribution is called the **canonical distribution**\n",
    "\n",
    "With a quadratic in $p$ and if $V(q) = \\frac{1}{2}q^2$ our distribution is gaussian and the level sets are ellipses of constant energy, as illustrated below, in a space called **phase space**, which is constructed by plotting the co-ordinates against the momenta: that is, it is just the space in which our augmented joint-distribution lives.\n",
    "\n",
    "![](../images/levelsets.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  Hamiltonian Mechanics\n",
    "\n",
    "The game now is to sample from this two-N-dimensional distribution and marginalize over the momenta to get the distribution from the $q$. To carry out this sampling, we'll use the physics equations of motion in the **Hamiltonian Formalism** (thus leading to the name Hamiltonian Monte Carlo) to \"glide\" over a level set.   Given a Hamiltonian H, the Hamiltonian equations of motion are as follows:\n",
    "\n",
    "$$\n",
    "\\begin{eqnarray}\n",
    "\\frac{dp}{dt}  &=& -\\frac {\\partial H}{\\partial q} \\\\\n",
    "\\frac{dq}{dt}  &=& \\frac {\\partial H}{\\partial p} \n",
    "\\end{eqnarray}\n",
    "$$\n",
    "\n",
    "If we have\n",
    "\n",
    "$$H=p^2/2m + V(q)$$\n",
    "\n",
    "then $\\frac{dp}{dt} = -\\frac {\\partial H}{\\partial q} = -\\frac {\\partial V}{\\partial q} = Force$ is newton's law, and the other equation sats that velocity is momentum divided by mass.\n",
    "\n",
    "### An example\n",
    "\n",
    "Here is an example of a harmonic oscillator with mass 1 and spring-constant 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr4AAAERCAYAAACdJaTTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x117bd6ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "min and max of H:  8.0 8.0\n"
     ]
    }
   ],
   "source": [
    "q_t = lambda t: 4. * np.cos(t)\n",
    "p_t = lambda t: -4. * np.sin(t)\n",
    "H_x_r_t = lambda t: q_t(t)**2/2. + p_t(t)**2/2.\n",
    "\n",
    "t = np.linspace(0,2*np.pi,100)\n",
    "\n",
    "plt.figure(figsize=[12,4])\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(t, q_t(t), label='q')\n",
    "plt.plot(t, p_t(t), label='p')\n",
    "plt.plot(t, H_x_r_t(t), label='H')\n",
    "plt.xlabel('t')\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(q_t(t), p_t(t), '.')\n",
    "plt.axis('equal')\n",
    "plt.xlabel('q(t)')\n",
    "plt.ylabel('p(t)')\n",
    "\n",
    "plt.show()\n",
    "\n",
    "print(\"min and max of H: \", np.min( H_x_r_t(t)), np.max( H_x_r_t(t)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The Hamiltonian is conserved\n",
    "\n",
    "Notice that in the example above the Hamiltonian is conserved. While the $q$ amd $p$ have time dependencies, the $H$ does not.\n",
    "\n",
    "If the Hamiltonian H doesn't have a functional dependence on time we see that \n",
    "\n",
    "$$ \\frac{dH}{dt} = \\sum_i \\left[\\frac {\\partial H}{\\partial q_i}\\frac{dq_i}{dt} + \\frac {\\partial H}{\\partial p_i}\\frac{dp_i}{dt} \\right] + \\frac{\\partial H}{\\partial t} $$\n",
    "\n",
    "Notice that I have explicitly indexed the co-ordinates and momenta to reflect the dimensionality of the prolem.\n",
    "\n",
    "Substituting in the Hamiltonian equations of motion we can see:\n",
    "$$ \\frac{dH}{dt} = \\sum_i \\left[\\frac {\\partial H}{\\partial q_i}\\frac {\\partial H}{\\partial p_i} + (\\frac {\\partial H}{\\partial p_i})(-\\frac {\\partial H}{\\partial q_i}) \\right] + \\frac{\\partial H}{\\partial t} $$\n",
    "\n",
    "\n",
    "So as long as the Hamiltonian $ H(\\{q_i, p_i\\}, t)$ depends only on $ \\{q_i, p_i \\} $ then $\\frac{\\partial H}{\\partial t} = 0$ and\n",
    "\n",
    "$$ \\frac{dH}{dt} = 0\\ $$\n",
    "\n",
    "and thus $ H(t + \\Delta t) = H(t) \\, \\forall t $.\n",
    "\n",
    "This time independence is crucial to reversibility: we cannot pick a direction of time in which this system is actually being run.\n",
    "\n",
    "### Reversibility\n",
    "\n",
    "Consider a transformation $T_s$ from $(q,p) \\to (q',p')$ where the primed quantities are just the variables at a \"later\" time $t' = t + s$. This mapping is 1-1 and thus has an inverse $T_{-s}$. This can be obtained by simply negating time in Hamilton's equations:\n",
    "\n",
    "$$\n",
    "\\begin{eqnarray}\n",
    "\\frac{dp}{d(-t)}  &=& -\\frac {\\partial H}{\\partial q} \\\\\n",
    "\\frac{dq}{d(-t)}  &=& \\frac {\\partial H}{\\partial p} \n",
    "\\end{eqnarray}\n",
    "$$\n",
    "\n",
    "This changes the equations of motion and would not be kosher, but we can\n",
    "preserve the equation, by merely transforming $p \\to -p$ and then \n",
    "\n",
    "$$\n",
    "\\begin{eqnarray}\n",
    "\\frac{d(-p)}{d(-t)}  &=& -\\frac {\\partial H}{\\partial q} \\\\\n",
    "\\frac{dq}{d(-t)}  &=& \\frac {\\partial H}{\\partial (-p)} \n",
    "\\end{eqnarray}\n",
    "$$\n",
    "\n",
    "and we have the old equations back. Notice that the momentum transformation does not change the Hamiltonian as long as the kinetic energy is symmetric in $p$.\n",
    "\n",
    "In other words this is what we need is this: to reverse $T_{s}$, flip the momentum, run Hamiltonian equations backwords in time until you get back to the original position and momentum in phase space at time t, and then flip the momentum again so it is pointing in the right direction (otherwise you would be going the wrong direction on re-running time forward...that is u want to make the tranformation which preserved the Hamiltonian equations above, twice.).\n",
    "\n",
    "In our oscillator example above, $T_{-s}$ is simply a rotation by $s$ radians counterclockwise in phase space, undoing the clockwise rotation created by $T_s$, and then fliiping the direction of rotation again.\n",
    "\n",
    "This is like in the superman movie! It will be critical in crafting a reversible Metropolis-Hastings proposal soon.\n",
    "\n",
    "### Volume in phase space is conserved\n",
    "\n",
    "Not only does no-explicit-time dynamics conserve the Hamiltonian, it also conserves a volume element in phase space. One can see this by If the determinant of the Jacobian of this transformation is 1, the volume is preserved.\n",
    "\n",
    "The transformation $T_{s}$ for small change $s = \\delta$ can be written as:\n",
    "\n",
    "$$ T_{\\delta} = \\left( \\begin{array}{c} q \\\\ p\\end{array} \\right) + \\delta \\left( \\begin{array}{c} \\frac{dq}{dt} \\\\ \\frac{dp}{dt}\\end{array} \\right)  + O(\\delta^2)$$\n",
    "\n",
    "The jacobian of this transformation is:\n",
    "$$\n",
    "\\left [\n",
    "\\begin{array}{c}\n",
    " 1 + \\delta \\frac{\\partial^2H}{\\partial q\\partial p} & \\delta \\frac{\\partial^2H}{\\partial p^2} \\\\\n",
    "\\delta \\frac{\\partial^2H}{\\partial q^2} &  1 - \\delta \\frac{\\partial^2H}{\\partial p\\partial q}\n",
    "\\end{array}\n",
    "\\right ]\n",
    "$$\n",
    "\n",
    "and thus the determinant is \n",
    "\n",
    "$$ 1 +  O(\\delta^2).$$\n",
    "\n",
    "Since we can compose the time transformation $s$ out of many such data, out Hamiltonian is volume preserving!\n",
    "\n",
    "Thus as our system evolves, any contraction or expansion in position space must be compensated by a respective expansion or compression in momentum space.\n",
    "\n",
    "As a result of this, the momenta we augment our distribution with must be **dual** to our pdf's parameters, transforming in the opposite way so that phase space volumes are invariant."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The microcanonical distribution\n",
    "\n",
    "We have expanded our pdf's parameter space by auxillary momenta. We have lift our target distribution in $q$ space to a joint probability distribution on phase space called the **canonical** distribution. In this phase space, we can explore the joint typical set by integrating Hamiltonian equations for a time.\n",
    "\n",
    "Since we will consider time-implicit Hamiltonians, these flows will be volume preserving and reversible, and indeed stick to a level surface of energy.\n",
    "\n",
    "Indeed we are traveling along constant $E$ or $p(p,q)$ foliates. This allows us to consider a related distribution to the canonical distribution, the **microcanonical distribution** (the first term in the product below). The canonical distribution can be written as a product of this microcanonical distribution and a **marginal energy distribution**:\n",
    "\n",
    "$$p(q,p) = p(\\theta_E \\vert E) p(E)$$\n",
    "\n",
    "where $\\theta_E$ indexes the position on the level set: ie a co-ordinate for example along the ellipses in our diagram above.\n",
    "\n",
    "The microcanonical distribution thus gives us the states for a given energy (or joint probability) while the marginal tells us the probability of a given level set in the typical set itself.\n",
    "\n",
    "This decomposition of the problem tells us how to explore the entire typical set. We integrate Hamilton's equations for a while to explore the microcanonical distribution on a given level set."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stochasticity\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But this then leaves us with the problem of having to go from one level-set to another: after all, we wish to explore the entire typical set. We thus need to stochastically explore the marginal energy distribution to go from one level-set to the other.\n",
    "\n",
    "### Sampling the momentum\n",
    "\n",
    "The idea for this once again comes from our mechanical satellite analog. Think of a rocket in a partivcular orbit firing its thrusters or letting momentum slip. It can achieve a different orbit.\n",
    "\n",
    "Thus the solution to exploring the marginal distribution is simple..after exploring a given level set for a while, we resample the momentum, and off to another level-set we go, as illustrated below:\n",
    "\n",
    "![](../images/momresample.png)\n",
    "\n",
    "Let $p(E \\vert q)$ as the transition distribution of energies induced by a momentum resampling using $p(p \\vert q) = - log\\, K(p,q)$ at a given position $q$. Our efficiency of this stochastic exploration over level sets then depends on how narrow this induced distribution is with respect to the marginal energy distribution:\n",
    "\n",
    "![](../images/lsetexp.png)\n",
    "\n",
    "If the transition distribution is narrow compared to the marginal energy distribution, then the random walk amongst level sets will proceed slowly. But if it matches the marginal energy distribution, we will generate nearly independent samples from the marginal energy distribution very efficiently.\n",
    "\n",
    "Thus we will draw $p$ from a distribution that is determined by the distribution of momentum, i.e. $p \\sim  N(0,\\sqrt{M})$ for example, and attempt to explore the level sets."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tuning\n",
    "\n",
    "It should be clear now that HMC needs to be tuned in at-least two ways:\n",
    "\n",
    "- what is our choice of kinetic energy and how does this impact the stochastic exploration of the marginal distribution\n",
    "- how long shall we integrate along a certain level set\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Choice of Kinetic Energy\n",
    "\n",
    "In theory we can choose any Kinetic Energy function K(p) that we deem useful.  The ideal kinetic energy would interact with the target distribution to make microcanonical exploration as easy and uniform as possible and marginal exploration well matched by the transition distribution.\n",
    "\n",
    "In practice we often use\n",
    "\n",
    "$$K(p) = p'M^{-1}p $$\n",
    "\n",
    "so that $K(p) = \\sum_i\\frac{p_i^2}{2m_i} $ and M is diagonal square matrix with elements $ 1/m_i $. This would allow us to particularize to various parameters.\n",
    "\n",
    "This is a euclidean-gaussian choice: euclidean as it corresponds to a euclidean norm on the co-ordinate space and gaussian as the conditional distribution it induces is Gaussian. Gaussian choices perform well in higher dimensions as marginal energy distributions become a convolution of more and more parameters and a kind of central-limit theorem holds.\n",
    "\n",
    "If one were to set the inverse mass matrix to the covariance of the target distribution then one would maximally decorrelate the target distribution. This can be seen thus: apply the transformation $p \\to \\sqrt{M^{-1}}p$ which simplifies the kinetic energy but transforms the potential via $q \\to \\sqrt{M}q$ due to the duality and volume preservation of $q,p$ phase space. Then you are effectively dividing $q$ by the squareoot of its distributions covariance and carrying out our decorrelation. \n",
    "\n",
    "We can do this by computing a target covariance in a warm-up (this is not burn-in) phase of our sampler and then setting the kinetic energy appropriately to make an optimized chain.\n",
    "\n",
    "\n",
    "### Choice of integration times\n",
    "\n",
    "Choosing a kinetic energy defines the shape of the level sets. Now the game is to glide along any one of these. How long should we do it? If we do it too short, we are not taking advantage of the coherent Hamiltonian updates and are back to our drunk random walk. If we do it too long, the topological compactness of these level sets in well defined problems will mean we revisit old points in phase space, yielding slower convergence.\n",
    "\n",
    "What we need to do is to find the point at which the orbital expectations converge to the spatial expectations..a sort of ergodicity. Integrating beyond this regime brings diminishing returns.\n",
    "\n",
    "This time can be represented with two parameters, $L$, the number of iterations for which we run the Hamiltonian dynamics, and $\\epsilon$ which is the (small) length of time each iteration is run. We say iterations as to solve Hamiltonian differential equations we will need to carry out a numerical solution of these equations with a small time-grid size. Then the total integration time is $L \\times \\epsilon$. These are parameters we need to play with, for this and other reasons (see below).\n",
    "\n",
    "But in general, no single optimal integration time will hold...this depends on which orbit in which level set we are. In general, for a distribution more heavy tailed than a Gaussian, as we move to higher energy level sets in the tails, we will need more time to integrate over there (this can be shown analytically in 1-D$). Thus if we used a static integration time everywhere, the tails might be very poorly explored (and presumably more of the typical set is in the tails of these distributuions than for thin tailed ones.\n",
    "\n",
    "Thus in general we will want to identify the integration time dynamically. This is the idea behind the No U-Turn Sampler (NUTS) which is the default in both pymc3 and Stan: the basic idea is that when the Hamiltonian flows starts to double back on itself and retrace its steps, we resample momentum again."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simulating Hamiltonian Systems:  Discretization\n",
    "\n",
    "Unfortunately we can't simulate Hamiltonian systems exactly and we have to rely on discretization!  What does that mean?  Well in the derivations above, we assumed that we could solve the differential equations in question exactly.  Since we're going to be sampling from our Hamiltonian dynamics we need to discretize our systems in to small timesteps of size $ \\epsilon $.  \n",
    "\n",
    "There are a number of discretization methods available for simulating dynamics systems.  It turns out that all of them aren't equally useful.  Some of the properties of the Let's start with the simplest, Euler discretization.  In all of our discretization examples we'll make the following assumptions:\n",
    "\n",
    "* the Hamiltonian has the form H(q, p) = U(q)+K(p)\n",
    "* K(p) = $p'M^{-1}p $ so that $K(p) = \\sum_i\\frac{p_i^2}{2m_i} $ and M is square with diagonal elements $ 1/m_i $\n",
    "\n",
    "Let's start with the most basic discretization, Euler's method.\n",
    "\n",
    "### Euler Discretization\n",
    "\n",
    "Euler discretization involves directly updating the momentum and position at each time step.  The algorithm is as follows:\n",
    "\n",
    "* $p_i(t + \\epsilon) = p_i(t) - \\epsilon \\frac{\\partial U}{\\partial q_i}\\mid_{q(t)} $\n",
    "* $q_i(t + \\epsilon) = q_i(t) + \\epsilon \\frac{p_i(t)}{m_i} $\n",
    "\n",
    "See code and plot below for harmonic oscillator, $ H = U(q) + K(p) = \\frac{q^2}{2} + \\frac{p^2}{2m} $.  Notice that the volume is not preserved as the  trajectory we plot for the harmonic oscillator, which should be an elipse, diverges to infinity and the Hamiltonian is not conserved as we'd expect it to be.  Something is wrong with this discretization.  \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x117f8dcf8>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtgAAAHwCAYAAACcxBjDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHQlsAADo6D+MIl1wFboQOAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x117f2ee48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "U = lambda q: q**2/2\n",
    "K = lambda p:  (p**2)/2\n",
    "dUdq= lambda q: q\n",
    "\n",
    "def euler(p0, q0, L, epsilon):\n",
    "    \n",
    "    ps = [p0]\n",
    "    qs = [q0]\n",
    "    \n",
    "    pprev = p0\n",
    "    qprev = q0\n",
    "    \n",
    "\n",
    "\n",
    "    # alternate full steps for position and momentum\n",
    "    for i in range(L):\n",
    "        p = pprev - epsilon*dUdq(qprev)\n",
    "        ps.append(p)\n",
    "        q = qprev + epsilon*pprev\n",
    "        qs.append(q)\n",
    "        \n",
    "        pprev = p\n",
    "        qprev = q\n",
    "\n",
    "\n",
    "    \n",
    "    return (ps, qs)\n",
    "\n",
    "\n",
    "eps,eqs = euler(1,0, 55, .3)\n",
    "h = [U(q) + K(p) for q, p in zip(eqs, eps)]\n",
    "\n",
    "plt.figure(figsize=[12,8])\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(eqs,eps, 'b.')\n",
    "plt.xlabel('q')\n",
    "plt.ylabel('p')\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(h)\n",
    "plt.ylabel('H')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Most ODE solvers, like the one we just wrote, suffer from drift:\n",
    "\n",
    "![](../images/nummodeerr.png)\n",
    "\n",
    "As we solve longer, error adds coherently, and our trajectory diverges from the true trajectory. This is because our discrete transformations like the Euler one above, do not preserve volume elements: the determinant of the Jacobian is larger than 1. The critical reason behind this is that we update both $p$ and $q$ simultaneously, accumulating error fast.\n",
    "\n",
    "If you think about the Jacobian, then the diagonal terms are 1. But if you update both $p$ and $q$ at once, the off-diagonals are non-zero for small but finite $\\epsilon$ (in other words u cant make $\\epsilon \\to 0$). Thus volume is not preserved.\n",
    "\n",
    "What we want are **symplectic** integrators, which preserve phase space volume elements."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Being symplectic: the Leapfrog method\n",
    "\n",
    "Any method that has only *shear* transforms: ie something that changes only one thing at a time, will work. The **leapfrog** method is one such idea, involving stage-wise updating the momentum and position.  The algorithm is as follows:\n",
    "\n",
    "* $p_i(t + \\frac{\\epsilon}{2}) = p_i(t) - \\frac{\\epsilon}{2}\\frac{\\partial U}{\\partial q_i}\\mid_{q(t)} $\n",
    "* $q_i(t + \\epsilon) = q_i(t) + \\epsilon \\frac{p_i(t+\\frac{\\epsilon}{2})}{m_i} $\n",
    "* $p_i(t + \\epsilon) = p_i(t+\\frac{\\epsilon}{2}) - \\frac{\\epsilon}{2}\\frac{\\partial U}{\\partial q_i}\\mid_{q(t+ \\epsilon)} $\n",
    "\n",
    "If we do this repeatedly, the first and last steps can be combined as you can see in the code below.\n",
    "\n",
    "Notice that  volume is preserved and the trajectory is very stable.  In addition the discretization is reversible.  The leapfrog method (named because updating momentum and position leapfrog each other) is a symplectic discretization.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "min and max of leapfrog H:  7.6190477193 8.42105060531\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtgAAAERCAYAAABB3NDkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1188595c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "U = lambda q: (q)**2/2.\n",
    "K = lambda p:  (p**2)/2.\n",
    "dUdq= lambda q: q\n",
    "\n",
    "def leapfrog(p0, q0, L, epsilon):\n",
    "    # keep all the epsilon steps\n",
    "    ps = [p0]\n",
    "    qs = [q0]\n",
    "    \n",
    "    # starting position\n",
    "    p = p0\n",
    "    q = q0\n",
    "    \n",
    "    # Make a half step for momentum at the beginning\n",
    "    p = p - epsilon*dUdq(q)/2.0\n",
    "\n",
    "    # alternate full steps for position and momentum\n",
    "    for i in range(L):\n",
    "        q = q + epsilon*p\n",
    "        qs.append(q)\n",
    "        if (i != L-1):\n",
    "            p = p - epsilon*dUdq(q)\n",
    "            ps.append(p)\n",
    "\n",
    "    #make a half step at the end\n",
    "    p = p - epsilon*dUdq(q)/2.\n",
    "    ps.append(p)\n",
    "\n",
    "    return (ps, qs)\n",
    "\n",
    "\n",
    "\n",
    "lps,lqs = leapfrog(4.,0., 1000, .1 )\n",
    "h = [U(q) + K(p) for q, p in zip(lqs, lps)]\n",
    "\n",
    "print(\"min and max of leapfrog H: \", np.min(h), np.max(h))\n",
    "\n",
    "plt.figure(figsize=[12,4])\n",
    "plt.subplot(1,2,1)\n",
    "plt.plot(lqs,lps, 'b.')\n",
    "plt.plot(x_t(t), r_t(t), 'r.')\n",
    "plt.xlabel('x(t)')\n",
    "plt.ylabel('y(t)')\n",
    "plt.axis('equal')\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.plot(h)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One thing to note, see the diagram and code above, is that even though the discretization is symplectic it (like the Euler modification) doesn't preserve the Hamiltonian perfectly.  That's because of the approximation errors incumbant in discretization.  The error is very stable and the value of the Hamiltonian at each step oscillates around the true value.  \n",
    "\n",
    "Because of this exact reversibility is lost. There is no reason to sweat though, superman to the rescue! When we sample from the potential in our HMC, we will use the acceptance probability to balance out the error in H."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The Acceptance Probability\n",
    "\n",
    "At the beginning of each sampling step we choose a new $p$ and our current $q$ and then run the Leapfrog algorithm for $L$ steps of size $ \\epsilon $. The new $q_L$ and $p_L$ at the end of our Leapfrog steps is our new proposed values, with our proposal being\n",
    "\n",
    "$$Q(q', p' \\vert q, p) = \\delta(q' - q_L) \\delta (p' - p_L).$$\n",
    "\n",
    "In other words, only propose something which gives us the integration sanctioned values..deterministic!\n",
    "\n",
    "This would seam to be the end of the story, but remember that even our symplectic integrator has errors. Errors make the Hamiltonian flow irreversible. In other words we can only propose states going forwards in time: the proposal probability for the time-reversed transition is 0 and the acceptance ratio undefined.\n",
    "\n",
    "But we know how to make the Hamiltonian transition reversible! Superman to the rescue. We augment the numerical integration with a momentum flip so that we now consider the transition:\n",
    "\n",
    "$$(q,p) \\to (q_L, -p_L)$$\n",
    "\n",
    "with proposal:\n",
    "\n",
    "$$Q(q', p' \\vert q, p) = \\delta(q' - q_L) \\delta (p' + p_L).$$\n",
    "\n",
    "This leads us to accept with probability \n",
    "\n",
    "$$A = \\min[1, \\frac{p(q_L, -p_L)\\delta(q_L -q_L)\\delta(-p_L + p_L)}{p(q,p)\\delta(q-q)\\delta(p-p)}] = \\min[1, \\frac{p(q_L, -p_L)}{p(q,p)}]$$\n",
    "\n",
    "and thus:\n",
    "\n",
    "$$A= \\min[1, \\exp(-U(q_L)+U(q)-K(p_L)+K(p)] $$\n",
    "\n",
    "This is exactly the kind of acceptance ratio we have seen in simulated annealing and metropolis, and stems from our definition of the joint(canonical) distribution.\n",
    "\n",
    "But the critical thing with HMC is that our **time evolution is on a level set**. So our acceptance probability is always close to 1, and we have a very efficient sampler. (You might have thought you could pick up a sample after each leapfrog update since it is coherent evolution of Hamiltonian equations, but then you would have to do a momentum reversal and acceptance comparision many many times).\n",
    "\n",
    "The momentum reversal could be left out if you are not within a more complex sampling scheme like HMC within gibbs since you will be resampling anyways. But if you are updating both a discrete parameter and a continuous parameter, you will want to reverse the sign so that you are using the correct $p$ when sampling from the conditionals."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](images/reversemom.png)\n",
    "\n",
    "In general we'll want to sum over all such points in the orbit, since we want time averages to represent a sample from the microcanonical distribution."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The HMC Algorithm\n",
    "\n",
    "Now that we have the tools in place, let's describe the HMC algorithm.  We're going to assume as above a Kinetic Energy $ K(p) = \\frac{p^{\\top}M^{-1}p}{2}  $.  The algorithm is as follows\n",
    "\n",
    "* for i=1:N_samples\n",
    "    1. Draw $  p \\sim N(0,M) $\n",
    "    2. Set $ q_{c} = q^{(i)} $ where the subscript $c$ stands for current\n",
    "    3. $ p_c = p $\n",
    "    4. Update momentum before going into LeapFrog stage: $ p^* = p_c - \\frac{\\epsilon * \\nabla U(q_c)}{2} $  \n",
    "    5. LeapFrog to get new proposals. For j=1:L\n",
    "        * $ q^{*} = q^{*} + \\epsilon  p $\n",
    "        * if not the last step, $ p = p - \\epsilon  \\nabla U(q) $\n",
    "    6. Complete leapfrog: $ p = p - \\frac{\\epsilon  \\nabla U(q)}{2} $\n",
    "    7. $ p^{*}= -p $\n",
    "    8. $ U_c = U(q_c), \\,\\,\\, K_c = \\frac{p_c^{\\top}M^{-1}p_c}{2} $\n",
    "    9. $ U^{*} = U(q^*), \\,\\,\\, K^{*} = \\frac{p^{\\top*}M^{-1}p^*}{2} $\n",
    "    10. $ r \\sim \\rm{Unif}(0,1) $\n",
    "    11. if $ r < e^{(U_{c}-U^{*}+K_{c}-K^{*})} $  \n",
    "       * accept $q_i = q^*$\n",
    "       * otherwise reject\n",
    "       "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### A simple implementation\n",
    "\n",
    "We implement the HMC algorithm below with a gaussian Kinetic energy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#constants\n",
    "\n",
    "def HMC(U,K,dUdq,N,q_0, p_0, epsilon=0.01, L=100):\n",
    "    p_mu=0.\n",
    "    p_sig=1.\n",
    "    \n",
    "    current_q = q_0\n",
    "    current_p = p_0\n",
    "    \n",
    "    H = np.zeros(N)\n",
    "    qall = np.zeros(N)\n",
    "    accept=0\n",
    "    for j in range(N):\n",
    "    \n",
    "        q = current_q\n",
    "        p = current_p\n",
    "        \n",
    "        #draw a new p\n",
    "        p = np.random.normal(p_mu, p_sig)\n",
    "        \n",
    "        current_p=p\n",
    "        \n",
    "        # leap frog\n",
    "        \n",
    "        # Make a half step for momentum at the beginning\n",
    "        p = p - epsilon*dUdq(q)/2.0\n",
    "        \n",
    "        \n",
    "        # alternate full steps for position and momentum\n",
    "        for i in range(L):\n",
    "            q = q + epsilon*p\n",
    "            if (i != L-1):\n",
    "                p = p - epsilon*dUdq(q)\n",
    "    \n",
    "        #make a half step at the end\n",
    "        p = p - epsilon*dUdq(q)/2.\n",
    "\n",
    "        # negate the momentum\n",
    "        p= -p;\n",
    "        current_U = U(current_q)\n",
    "        current_K = K(current_p)\n",
    "\n",
    "        proposed_U = U(q)\n",
    "        proposed_K = K(p)\n",
    "        A=np.exp( current_U-proposed_U+current_K-proposed_K)\n",
    "    \n",
    "        # accept/reject\n",
    "        if np.random.rand() < A:\n",
    "            current_q = q\n",
    "            qall[j]=q\n",
    "            accept+=1\n",
    "        else:\n",
    "            qall[j] = current_q\n",
    "        \n",
    "        H[j] = U(current_q)+K(current_p)\n",
    "    print(\"accept=\",accept/np.double(N))\n",
    "    return H, qall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "# functions\n",
    "U = lambda q: q**2/2.\n",
    "K = lambda p:  (p**2)/2.\n",
    "dUdq= lambda q: q\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accept= 1.0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n7v8DQpXjhp3AHVQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118d060f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "H, qall= HMC(U=U,K=K,dUdq=dUdq,N=10000,q_0=0, p_0=-4, epsilon=0.01, L=200)\n",
    "plt.hist(qall, bins=50, normed=True)\n",
    "x = np.linspace(-4,4,100)\n",
    "plt.plot(x, sp.stats.norm.pdf(x),'r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We compare it to a MH sampler with the same number of steps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "def MH_simple(p, n, sig, x0):\n",
    "    x_prev = x0\n",
    "    x=[]\n",
    "    k=1\n",
    "    i=0\n",
    "    while i<n:\n",
    "        x_star = np.random.normal(x_prev, sig)\n",
    "        P_star = p(x_star)\n",
    "        P_prev = p(x_prev)\n",
    "        U =  np.random.uniform()\n",
    "        \n",
    "        A =  P_star/P_prev\n",
    "        if U < A:\n",
    "            x.append(x_star)\n",
    "            i = i + 1\n",
    "            x_prev = x_star\n",
    "            k +=1\n",
    "        else :\n",
    "            x.append(x_prev)\n",
    "            x_prev = x[i]  \n",
    "            i = i + 1\n",
    "            \n",
    "    print(\"accept=\",k/np.double(n))\n",
    "    return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accept= 0.3002\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAECCAYAAADw0Rw8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFCXFqUSvKIqS4v5/8PPEauAn1WgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118cfe470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "samples_mh = MH_simple(p=P, n=10000, sig=4.0, x0=0)\n",
    "\n",
    "plt.subplot(1,2,1)\n",
    "plt.hist(qall, bins=50, normed=True)\n",
    "x = np.linspace(-4,4,100)\n",
    "plt.plot(x, sp.stats.norm.pdf(x),'r')\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.hist(samples_mh, bins=50, normed=True)\n",
    "x = np.linspace(-4,4,100)\n",
    "plt.plot(x, sp.stats.norm.pdf(x),'r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here we see that the MH acceptance ration is much lower and the correlation much higher!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEKCAYAAAAb7IIBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKsoASFJRBkCSivo/39RLPuELPf4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1192dc588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def corrplot(trace,  maxlags=100):\n",
    "    plt.acorr(trace-np.mean(trace),  normed=True, maxlags=maxlags);\n",
    "    plt.xlim([0, maxlags])\n",
    "corrplot(qall)\n",
    "plt.title('hmc');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEKCAYAAAAb7IIBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119202f98>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "corrplot(samples_mh)\n",
    "plt.title('mh');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}