diff options
Diffstat (limited to 'notebooks/plot_gromov_barycenter.ipynb')
| -rw-r--r-- | notebooks/plot_gromov_barycenter.ipynb | 471 |
1 files changed, 247 insertions, 224 deletions
diff --git a/notebooks/plot_gromov_barycenter.ipynb b/notebooks/plot_gromov_barycenter.ipynb index 8102bcf..c931a6c 100644 --- a/notebooks/plot_gromov_barycenter.ipynb +++ b/notebooks/plot_gromov_barycenter.ipynb @@ -32,20 +32,20 @@ }, "outputs": [], "source": [ - "# Author: Erwan Vautier <erwan.vautier@gmail.com>\r\n", - "# Nicolas Courty <ncourty@irisa.fr>\r\n", - "#\r\n", - "# License: MIT License\r\n", - "\r\n", - "\r\n", - "import numpy as np\r\n", - "import scipy as sp\r\n", - "\r\n", - "import scipy.ndimage as spi\r\n", - "import matplotlib.pylab as pl\r\n", - "from sklearn import manifold\r\n", - "from sklearn.decomposition import PCA\r\n", - "\r\n", + "# Author: Erwan Vautier <erwan.vautier@gmail.com>\n", + "# Nicolas Courty <ncourty@irisa.fr>\n", + "#\n", + "# License: MIT License\n", + "\n", + "\n", + "import numpy as np\n", + "import scipy as sp\n", + "\n", + "import scipy.ndimage as spi\n", + "import matplotlib.pylab as pl\n", + "from sklearn import manifold\n", + "from sklearn.decomposition import PCA\n", + "\n", "import ot" ] }, @@ -53,11 +53,11 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Smacof MDS\r\n", - " ----------\r\n", - "\r\n", - " This function allows to find an embedding of points given a dissimilarity matrix\r\n", - " that will be given by the output of the algorithm\r\n", + "Smacof MDS\n", + "----------\n", + "\n", + "This function allows to find an embedding of points given a dissimilarity matrix\n", + "that will be given by the output of the algorithm\n", "\n" ] }, @@ -69,49 +69,49 @@ }, "outputs": [], "source": [ - "def smacof_mds(C, dim, max_iter=3000, eps=1e-9):\r\n", - " \"\"\"\r\n", - " Returns an interpolated point cloud following the dissimilarity matrix C\r\n", - " using SMACOF multidimensional scaling (MDS) in specific dimensionned\r\n", - " target space\r\n", - "\r\n", - " Parameters\r\n", - " ----------\r\n", - " C : ndarray, shape (ns, ns)\r\n", - " dissimilarity matrix\r\n", - " dim : int\r\n", - " dimension of the targeted space\r\n", - " max_iter : int\r\n", - " Maximum number of iterations of the SMACOF algorithm for a single run\r\n", - " eps : float\r\n", - " relative tolerance w.r.t stress to declare converge\r\n", - "\r\n", - " Returns\r\n", - " -------\r\n", - " npos : ndarray, shape (R, dim)\r\n", - " Embedded coordinates of the interpolated point cloud (defined with\r\n", - " one isometry)\r\n", - " \"\"\"\r\n", - "\r\n", - " rng = np.random.RandomState(seed=3)\r\n", - "\r\n", - " mds = manifold.MDS(\r\n", - " dim,\r\n", - " max_iter=max_iter,\r\n", - " eps=1e-9,\r\n", - " dissimilarity='precomputed',\r\n", - " n_init=1)\r\n", - " pos = mds.fit(C).embedding_\r\n", - "\r\n", - " nmds = manifold.MDS(\r\n", - " 2,\r\n", - " max_iter=max_iter,\r\n", - " eps=1e-9,\r\n", - " dissimilarity=\"precomputed\",\r\n", - " random_state=rng,\r\n", - " n_init=1)\r\n", - " npos = nmds.fit_transform(C, init=pos)\r\n", - "\r\n", + "def smacof_mds(C, dim, max_iter=3000, eps=1e-9):\n", + " \"\"\"\n", + " Returns an interpolated point cloud following the dissimilarity matrix C\n", + " using SMACOF multidimensional scaling (MDS) in specific dimensionned\n", + " target space\n", + "\n", + " Parameters\n", + " ----------\n", + " C : ndarray, shape (ns, ns)\n", + " dissimilarity matrix\n", + " dim : int\n", + " dimension of the targeted space\n", + " max_iter : int\n", + " Maximum number of iterations of the SMACOF algorithm for a single run\n", + " eps : float\n", + " relative tolerance w.r.t stress to declare converge\n", + "\n", + " Returns\n", + " -------\n", + " npos : ndarray, shape (R, dim)\n", + " Embedded coordinates of the interpolated point cloud (defined with\n", + " one isometry)\n", + " \"\"\"\n", + "\n", + " rng = np.random.RandomState(seed=3)\n", + "\n", + " mds = manifold.MDS(\n", + " dim,\n", + " max_iter=max_iter,\n", + " eps=1e-9,\n", + " dissimilarity='precomputed',\n", + " n_init=1)\n", + " pos = mds.fit(C).embedding_\n", + "\n", + " nmds = manifold.MDS(\n", + " 2,\n", + " max_iter=max_iter,\n", + " eps=1e-9,\n", + " dissimilarity=\"precomputed\",\n", + " random_state=rng,\n", + " n_init=1)\n", + " npos = nmds.fit_transform(C, init=pos)\n", + "\n", " return npos" ] }, @@ -119,10 +119,10 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Data preparation\r\n", - " ----------------\r\n", - "\r\n", - " The four distributions are constructed from 4 simple images\r\n", + "Data preparation\n", + "----------------\n", + "\n", + "The four distributions are constructed from 4 simple images\n", "\n" ] }, @@ -132,31 +132,54 @@ "metadata": { "collapsed": false }, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/home/rflamary/.local/lib/python3.5/site-packages/ipykernel_launcher.py:6: DeprecationWarning: `imread` is deprecated!\n", + "`imread` is deprecated in SciPy 1.0.0.\n", + "Use ``matplotlib.pyplot.imread`` instead.\n", + " \n", + "/home/rflamary/.local/lib/python3.5/site-packages/ipykernel_launcher.py:7: DeprecationWarning: `imread` is deprecated!\n", + "`imread` is deprecated in SciPy 1.0.0.\n", + "Use ``matplotlib.pyplot.imread`` instead.\n", + " import sys\n", + "/home/rflamary/.local/lib/python3.5/site-packages/ipykernel_launcher.py:8: DeprecationWarning: `imread` is deprecated!\n", + "`imread` is deprecated in SciPy 1.0.0.\n", + "Use ``matplotlib.pyplot.imread`` instead.\n", + " \n", + "/home/rflamary/.local/lib/python3.5/site-packages/ipykernel_launcher.py:9: DeprecationWarning: `imread` is deprecated!\n", + "`imread` is deprecated in SciPy 1.0.0.\n", + "Use ``matplotlib.pyplot.imread`` instead.\n", + " if __name__ == '__main__':\n" + ] + } + ], "source": [ - "def im2mat(I):\r\n", - " \"\"\"Converts and image to matrix (one pixel per line)\"\"\"\r\n", - " return I.reshape((I.shape[0] * I.shape[1], I.shape[2]))\r\n", - "\r\n", - "\r\n", - "square = spi.imread('../data/square.png').astype(np.float64)[:, :, 2] / 256\r\n", - "cross = spi.imread('../data/cross.png').astype(np.float64)[:, :, 2] / 256\r\n", - "triangle = spi.imread('../data/triangle.png').astype(np.float64)[:, :, 2] / 256\r\n", - "star = spi.imread('../data/star.png').astype(np.float64)[:, :, 2] / 256\r\n", - "\r\n", - "shapes = [square, cross, triangle, star]\r\n", - "\r\n", - "S = 4\r\n", - "xs = [[] for i in range(S)]\r\n", - "\r\n", - "\r\n", - "for nb in range(4):\r\n", - " for i in range(8):\r\n", - " for j in range(8):\r\n", - " if shapes[nb][i, j] < 0.95:\r\n", - " xs[nb].append([j, 8 - i])\r\n", - "\r\n", - "xs = np.array([np.array(xs[0]), np.array(xs[1]),\r\n", + "def im2mat(I):\n", + " \"\"\"Converts and image to matrix (one pixel per line)\"\"\"\n", + " return I.reshape((I.shape[0] * I.shape[1], I.shape[2]))\n", + "\n", + "\n", + "square = spi.imread('../data/square.png').astype(np.float64)[:, :, 2] / 256\n", + "cross = spi.imread('../data/cross.png').astype(np.float64)[:, :, 2] / 256\n", + "triangle = spi.imread('../data/triangle.png').astype(np.float64)[:, :, 2] / 256\n", + "star = spi.imread('../data/star.png').astype(np.float64)[:, :, 2] / 256\n", + "\n", + "shapes = [square, cross, triangle, star]\n", + "\n", + "S = 4\n", + "xs = [[] for i in range(S)]\n", + "\n", + "\n", + "for nb in range(4):\n", + " for i in range(8):\n", + " for j in range(8):\n", + " if shapes[nb][i, j] < 0.95:\n", + " xs[nb].append([j, 8 - i])\n", + "\n", + "xs = np.array([np.array(xs[0]), np.array(xs[1]),\n", " np.array(xs[2]), np.array(xs[3])])" ] }, @@ -164,8 +187,8 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Barycenter computation\r\n", - "----------------------\r\n", + "Barycenter computation\n", + "----------------------\n", "\n" ] }, @@ -177,45 +200,45 @@ }, "outputs": [], "source": [ - "ns = [len(xs[s]) for s in range(S)]\r\n", - "n_samples = 30\r\n", - "\r\n", - "\"\"\"Compute all distances matrices for the four shapes\"\"\"\r\n", - "Cs = [sp.spatial.distance.cdist(xs[s], xs[s]) for s in range(S)]\r\n", - "Cs = [cs / cs.max() for cs in Cs]\r\n", - "\r\n", - "ps = [ot.unif(ns[s]) for s in range(S)]\r\n", - "p = ot.unif(n_samples)\r\n", - "\r\n", - "\r\n", - "lambdast = [[float(i) / 3, float(3 - i) / 3] for i in [1, 2]]\r\n", - "\r\n", - "Ct01 = [0 for i in range(2)]\r\n", - "for i in range(2):\r\n", - " Ct01[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[0], Cs[1]],\r\n", - " [ps[0], ps[1]\r\n", - " ], p, lambdast[i], 'square_loss', 5e-4,\r\n", - " max_iter=100, tol=1e-3)\r\n", - "\r\n", - "Ct02 = [0 for i in range(2)]\r\n", - "for i in range(2):\r\n", - " Ct02[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[0], Cs[2]],\r\n", - " [ps[0], ps[2]\r\n", - " ], p, lambdast[i], 'square_loss', 5e-4,\r\n", - " max_iter=100, tol=1e-3)\r\n", - "\r\n", - "Ct13 = [0 for i in range(2)]\r\n", - "for i in range(2):\r\n", - " Ct13[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[1], Cs[3]],\r\n", - " [ps[1], ps[3]\r\n", - " ], p, lambdast[i], 'square_loss', 5e-4,\r\n", - " max_iter=100, tol=1e-3)\r\n", - "\r\n", - "Ct23 = [0 for i in range(2)]\r\n", - "for i in range(2):\r\n", - " Ct23[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[2], Cs[3]],\r\n", - " [ps[2], ps[3]\r\n", - " ], p, lambdast[i], 'square_loss', 5e-4,\r\n", + "ns = [len(xs[s]) for s in range(S)]\n", + "n_samples = 30\n", + "\n", + "\"\"\"Compute all distances matrices for the four shapes\"\"\"\n", + "Cs = [sp.spatial.distance.cdist(xs[s], xs[s]) for s in range(S)]\n", + "Cs = [cs / cs.max() for cs in Cs]\n", + "\n", + "ps = [ot.unif(ns[s]) for s in range(S)]\n", + "p = ot.unif(n_samples)\n", + "\n", + "\n", + "lambdast = [[float(i) / 3, float(3 - i) / 3] for i in [1, 2]]\n", + "\n", + "Ct01 = [0 for i in range(2)]\n", + "for i in range(2):\n", + " Ct01[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[0], Cs[1]],\n", + " [ps[0], ps[1]\n", + " ], p, lambdast[i], 'square_loss', # 5e-4,\n", + " max_iter=100, tol=1e-3)\n", + "\n", + "Ct02 = [0 for i in range(2)]\n", + "for i in range(2):\n", + " Ct02[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[0], Cs[2]],\n", + " [ps[0], ps[2]\n", + " ], p, lambdast[i], 'square_loss', # 5e-4,\n", + " max_iter=100, tol=1e-3)\n", + "\n", + "Ct13 = [0 for i in range(2)]\n", + "for i in range(2):\n", + " Ct13[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[1], Cs[3]],\n", + " [ps[1], ps[3]\n", + " ], p, lambdast[i], 'square_loss', # 5e-4,\n", + " max_iter=100, tol=1e-3)\n", + "\n", + "Ct23 = [0 for i in range(2)]\n", + "for i in range(2):\n", + " Ct23[i] = ot.gromov.gromov_barycenters(n_samples, [Cs[2], Cs[3]],\n", + " [ps[2], ps[3]\n", + " ], p, lambdast[i], 'square_loss', # 5e-4,\n", " max_iter=100, tol=1e-3)" ] }, @@ -223,10 +246,10 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Visualization\r\n", - " -------------\r\n", - "\r\n", - " The PCA helps in getting consistency between the rotations\r\n", + "Visualization\n", + "-------------\n", + "\n", + "The PCA helps in getting consistency between the rotations\n", "\n" ] }, @@ -240,7 +263,7 @@ { "data": { "text/plain": [ - "<matplotlib.collections.PathCollection at 0x7fd293df1e50>" + "<matplotlib.collections.PathCollection at 0x7fcec1ef00b8>" ] }, "execution_count": 6, @@ -249,9 +272,9 @@ }, { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmMAAAJDCAYAAABHZBNLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nKWMAABMpYwAAE30Dk13yOM0vHUgAAAAASUVORK5CYII=\n", + "image/png": "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\n", "text/plain": [ - "<matplotlib.figure.Figure at 0x7fd294476ed0>" + "<matplotlib.figure.Figure at 0x7fcec20be470>" ] }, "metadata": {}, @@ -259,108 +282,108 @@ } ], "source": [ - "clf = PCA(n_components=2)\r\n", - "npos = [0, 0, 0, 0]\r\n", - "npos = [smacof_mds(Cs[s], 2) for s in range(S)]\r\n", - "\r\n", - "npost01 = [0, 0]\r\n", - "npost01 = [smacof_mds(Ct01[s], 2) for s in range(2)]\r\n", - "npost01 = [clf.fit_transform(npost01[s]) for s in range(2)]\r\n", - "\r\n", - "npost02 = [0, 0]\r\n", - "npost02 = [smacof_mds(Ct02[s], 2) for s in range(2)]\r\n", - "npost02 = [clf.fit_transform(npost02[s]) for s in range(2)]\r\n", - "\r\n", - "npost13 = [0, 0]\r\n", - "npost13 = [smacof_mds(Ct13[s], 2) for s in range(2)]\r\n", - "npost13 = [clf.fit_transform(npost13[s]) for s in range(2)]\r\n", - "\r\n", - "npost23 = [0, 0]\r\n", - "npost23 = [smacof_mds(Ct23[s], 2) for s in range(2)]\r\n", - "npost23 = [clf.fit_transform(npost23[s]) for s in range(2)]\r\n", - "\r\n", - "\r\n", - "fig = pl.figure(figsize=(10, 10))\r\n", - "\r\n", - "ax1 = pl.subplot2grid((4, 4), (0, 0))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax1.scatter(npos[0][:, 0], npos[0][:, 1], color='r')\r\n", - "\r\n", - "ax2 = pl.subplot2grid((4, 4), (0, 1))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax2.scatter(npost01[1][:, 0], npost01[1][:, 1], color='b')\r\n", - "\r\n", - "ax3 = pl.subplot2grid((4, 4), (0, 2))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax3.scatter(npost01[0][:, 0], npost01[0][:, 1], color='b')\r\n", - "\r\n", - "ax4 = pl.subplot2grid((4, 4), (0, 3))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax4.scatter(npos[1][:, 0], npos[1][:, 1], color='r')\r\n", - "\r\n", - "ax5 = pl.subplot2grid((4, 4), (1, 0))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax5.scatter(npost02[1][:, 0], npost02[1][:, 1], color='b')\r\n", - "\r\n", - "ax6 = pl.subplot2grid((4, 4), (1, 3))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax6.scatter(npost13[1][:, 0], npost13[1][:, 1], color='b')\r\n", - "\r\n", - "ax7 = pl.subplot2grid((4, 4), (2, 0))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax7.scatter(npost02[0][:, 0], npost02[0][:, 1], color='b')\r\n", - "\r\n", - "ax8 = pl.subplot2grid((4, 4), (2, 3))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax8.scatter(npost13[0][:, 0], npost13[0][:, 1], color='b')\r\n", - "\r\n", - "ax9 = pl.subplot2grid((4, 4), (3, 0))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax9.scatter(npos[2][:, 0], npos[2][:, 1], color='r')\r\n", - "\r\n", - "ax10 = pl.subplot2grid((4, 4), (3, 1))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax10.scatter(npost23[1][:, 0], npost23[1][:, 1], color='b')\r\n", - "\r\n", - "ax11 = pl.subplot2grid((4, 4), (3, 2))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", - "ax11.scatter(npost23[0][:, 0], npost23[0][:, 1], color='b')\r\n", - "\r\n", - "ax12 = pl.subplot2grid((4, 4), (3, 3))\r\n", - "pl.xlim((-1, 1))\r\n", - "pl.ylim((-1, 1))\r\n", + "clf = PCA(n_components=2)\n", + "npos = [0, 0, 0, 0]\n", + "npos = [smacof_mds(Cs[s], 2) for s in range(S)]\n", + "\n", + "npost01 = [0, 0]\n", + "npost01 = [smacof_mds(Ct01[s], 2) for s in range(2)]\n", + "npost01 = [clf.fit_transform(npost01[s]) for s in range(2)]\n", + "\n", + "npost02 = [0, 0]\n", + "npost02 = [smacof_mds(Ct02[s], 2) for s in range(2)]\n", + "npost02 = [clf.fit_transform(npost02[s]) for s in range(2)]\n", + "\n", + "npost13 = [0, 0]\n", + "npost13 = [smacof_mds(Ct13[s], 2) for s in range(2)]\n", + "npost13 = [clf.fit_transform(npost13[s]) for s in range(2)]\n", + "\n", + "npost23 = [0, 0]\n", + "npost23 = [smacof_mds(Ct23[s], 2) for s in range(2)]\n", + "npost23 = [clf.fit_transform(npost23[s]) for s in range(2)]\n", + "\n", + "\n", + "fig = pl.figure(figsize=(10, 10))\n", + "\n", + "ax1 = pl.subplot2grid((4, 4), (0, 0))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax1.scatter(npos[0][:, 0], npos[0][:, 1], color='r')\n", + "\n", + "ax2 = pl.subplot2grid((4, 4), (0, 1))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax2.scatter(npost01[1][:, 0], npost01[1][:, 1], color='b')\n", + "\n", + "ax3 = pl.subplot2grid((4, 4), (0, 2))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax3.scatter(npost01[0][:, 0], npost01[0][:, 1], color='b')\n", + "\n", + "ax4 = pl.subplot2grid((4, 4), (0, 3))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax4.scatter(npos[1][:, 0], npos[1][:, 1], color='r')\n", + "\n", + "ax5 = pl.subplot2grid((4, 4), (1, 0))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax5.scatter(npost02[1][:, 0], npost02[1][:, 1], color='b')\n", + "\n", + "ax6 = pl.subplot2grid((4, 4), (1, 3))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax6.scatter(npost13[1][:, 0], npost13[1][:, 1], color='b')\n", + "\n", + "ax7 = pl.subplot2grid((4, 4), (2, 0))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax7.scatter(npost02[0][:, 0], npost02[0][:, 1], color='b')\n", + "\n", + "ax8 = pl.subplot2grid((4, 4), (2, 3))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax8.scatter(npost13[0][:, 0], npost13[0][:, 1], color='b')\n", + "\n", + "ax9 = pl.subplot2grid((4, 4), (3, 0))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax9.scatter(npos[2][:, 0], npos[2][:, 1], color='r')\n", + "\n", + "ax10 = pl.subplot2grid((4, 4), (3, 1))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax10.scatter(npost23[1][:, 0], npost23[1][:, 1], color='b')\n", + "\n", + "ax11 = pl.subplot2grid((4, 4), (3, 2))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", + "ax11.scatter(npost23[0][:, 0], npost23[0][:, 1], color='b')\n", + "\n", + "ax12 = pl.subplot2grid((4, 4), (3, 3))\n", + "pl.xlim((-1, 1))\n", + "pl.ylim((-1, 1))\n", "ax12.scatter(npos[3][:, 0], npos[3][:, 1], color='r')" ] } ], "metadata": { "kernelspec": { - "display_name": "Python 2", + "display_name": "Python 3", "language": "python", - "name": "python2" + "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", - "version": 2 + "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", - "pygments_lexer": "ipython2", - "version": "2.7.12" + "pygments_lexer": "ipython3", + "version": "3.5.2" } }, "nbformat": 4, |