SOM_with_Somoclu.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"SOM_with_Somoclu.ipynb","version":"0.3.2","provenance":[]},"kernelspec":{"display_name":"Python 3.5","language":"python","name":"python3"}},"cells":[{"metadata":{"id":"IYVxCZALTL19","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":575},"outputId":"3a7c2709-5ff2-46af-97f8-e35d5905993d","executionInfo":{"status":"ok","timestamp":1543827728944,"user_tz":-120,"elapsed":47921,"user":{"displayName":"Νtdark 13","photoUrl":"","userId":"09958224867973618888"}}},"cell_type":"code","source":["!pip install --upgrade pip\n","!pip install --upgrade numpy\n","!pip install --upgrade pandas\n","!pip install --upgrade scikit-learn"],"execution_count":1,"outputs":[{"output_type":"stream","text":["Requirement already up-to-date: pip in /usr/local/lib/python3.6/dist-packages (18.1)\n","Collecting numpy\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/ff/7f/9d804d2348471c67a7d8b5f84f9bc59fd1cefa148986f2b74552f8573555/numpy-1.15.4-cp36-cp36m-manylinux1_x86_64.whl (13.9MB)\n","\u001b[K    100% |████████████████████████████████| 13.9MB 2.3MB/s \n","\u001b[?25hInstalling collected packages: numpy\n","  Found existing installation: numpy 1.14.6\n","    Uninstalling numpy-1.14.6:\n","      Successfully uninstalled numpy-1.14.6\n","Successfully installed numpy-1.15.4\n","Collecting pandas\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/e1/d8/feeb346d41f181e83fba45224ab14a8d8af019b48af742e047f3845d8cff/pandas-0.23.4-cp36-cp36m-manylinux1_x86_64.whl (8.9MB)\n","\u001b[K    100% |████████████████████████████████| 8.9MB 4.0MB/s \n","\u001b[?25hRequirement already satisfied, skipping upgrade: numpy>=1.9.0 in /usr/local/lib/python3.6/dist-packages (from pandas) (1.15.4)\n","Requirement already satisfied, skipping upgrade: pytz>=2011k in /usr/local/lib/python3.6/dist-packages (from pandas) (2018.7)\n","Requirement already satisfied, skipping upgrade: python-dateutil>=2.5.0 in /usr/local/lib/python3.6/dist-packages (from pandas) (2.5.3)\n","Requirement already satisfied, skipping upgrade: six>=1.5 in /usr/local/lib/python3.6/dist-packages (from python-dateutil>=2.5.0->pandas) (1.11.0)\n","Installing collected packages: pandas\n","  Found existing installation: pandas 0.22.0\n","    Uninstalling pandas-0.22.0:\n","      Successfully uninstalled pandas-0.22.0\n","Successfully installed pandas-0.23.4\n","Collecting scikit-learn\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/10/26/d04320c3edf2d59b1fcd0720b46753d4d603a76e68d8ad10a9b92ab06db2/scikit_learn-0.20.1-cp36-cp36m-manylinux1_x86_64.whl (5.4MB)\n","\u001b[K    100% |████████████████████████████████| 5.4MB 7.4MB/s \n","\u001b[?25hRequirement already satisfied, skipping upgrade: numpy>=1.8.2 in /usr/local/lib/python3.6/dist-packages (from scikit-learn) (1.15.4)\n","Requirement already satisfied, skipping upgrade: scipy>=0.13.3 in /usr/local/lib/python3.6/dist-packages (from scikit-learn) (1.1.0)\n","Installing collected packages: scikit-learn\n","  Found existing installation: scikit-learn 0.19.2\n","    Uninstalling scikit-learn-0.19.2:\n","      Successfully uninstalled scikit-learn-0.19.2\n","Successfully installed scikit-learn-0.20.1\n"],"name":"stdout"}]},{"metadata":{"id":"k-D8dzoxTL2M","colab_type":"text"},"cell_type":"markdown","source":["# SOM with Somoclu\n","\n","Στο notebook αυτό θα χρησιμοποιήσουμε τη βιβλιοθήκη [Somoclu](https://somoclu.readthedocs.io/en/stable/index.html). Πρόκειται για μια σύγχρονη και βελτιστοποιημένη υλοποίηση των SOM, παραλληλοποιημένη για CPU και GPU. Πρώτα την εγκαθιστούμε:\n"]},{"metadata":{"id":"KLuyf03CTL2P","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":332},"outputId":"579a0578-4582-4ce1-a6f0-52d1dda2536d","executionInfo":{"status":"ok","timestamp":1543827771201,"user_tz":-120,"elapsed":9792,"user":{"displayName":"Νtdark 13","photoUrl":"","userId":"09958224867973618888"}}},"cell_type":"code","source":["!pip install --upgrade somoclu"],"execution_count":2,"outputs":[{"output_type":"stream","text":["Collecting somoclu\n","\u001b[?25l  Downloading https://files.pythonhosted.org/packages/15/83/4b76845d8a2b0d70c27e7bdad363f36187b2adfc2851724ae76b47e0e348/somoclu-1.7.5.tar.gz (6.1MB)\n","\u001b[K    100% |████████████████████████████████| 6.2MB 5.1MB/s \n","\u001b[?25hRequirement already satisfied, skipping upgrade: numpy in /usr/local/lib/python3.6/dist-packages (from somoclu) (1.15.4)\n","Requirement already satisfied, skipping upgrade: matplotlib in /usr/local/lib/python3.6/dist-packages (from somoclu) (2.1.2)\n","Requirement already satisfied, skipping upgrade: scipy in /usr/local/lib/python3.6/dist-packages (from somoclu) (1.1.0)\n","Requirement already satisfied, skipping upgrade: pytz in /usr/local/lib/python3.6/dist-packages (from matplotlib->somoclu) (2018.7)\n","Requirement already satisfied, skipping upgrade: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->somoclu) (2.5.3)\n","Requirement already satisfied, skipping upgrade: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib->somoclu) (0.10.0)\n","Requirement already satisfied, skipping upgrade: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib->somoclu) (2.3.0)\n","Requirement already satisfied, skipping upgrade: six>=1.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib->somoclu) (1.11.0)\n","Building wheels for collected packages: somoclu\n","  Running setup.py bdist_wheel for somoclu ... \u001b[?25l-\b \b\\\b \b|\b \b/\b \b-\b \b\\\b \b|\b \b/\b \bdone\n","\u001b[?25h  Stored in directory: /root/.cache/pip/wheels/85/3c/e5/5b1c535e93453367a8e3d35a377d3c668da00369b1a4302ad8\n","Successfully built somoclu\n","Installing collected packages: somoclu\n","Successfully installed somoclu-1.7.5\n"],"name":"stdout"}]},{"metadata":{"id":"_kfCHqWpTL2Y","colab_type":"text"},"cell_type":"markdown","source":["## Εισαγωγή βιβλιοθηκών\n","\n","Εισάγουμε numpy, matplotlib και mpl_toolkits, και somoclu:"]},{"metadata":{"id":"p5MnZYh7TL2c","colab_type":"code","colab":{}},"cell_type":"code","source":["import numpy as np\n","import matplotlib.pyplot as plt\n","from mpl_toolkits.mplot3d import Axes3D\n","import somoclu\n","%matplotlib inline"],"execution_count":0,"outputs":[]},{"metadata":{"id":"-Zv-4_66TL2l","colab_type":"text"},"cell_type":"markdown","source":["## Δημιουργία και οπτικοποίηση dataset"]},{"metadata":{"id":"df0ai09uTL2n","colab_type":"text"},"cell_type":"markdown","source":["Δημιουργούμε 3 τεχνητές ομάδες 3D δεδομένων και εμφανίζουμε το γράφημα αναπαράστασής τους με διαφορετικό χρώμα για κάθε ομάδα:"]},{"metadata":{"id":"Y_sWg-2tTL2q","colab_type":"code","colab":{"base_uri":"https://localhost:8080/","height":319},"outputId":"25987dc3-cd29-4821-8403-c65f3dfd1b0a","executionInfo":{"status":"ok","timestamp":1543827784532,"user_tz":-120,"elapsed":932,"user":{"displayName":"Νtdark 13","photoUrl":"","userId":"09958224867973618888"}}},"cell_type":"code","source":["samples_in_each_class = 200\n","c1 = np.random.rand(samples_in_each_class, 3)/5\n","c2 = (0.6, 0.1, 0.05) + np.random.rand(samples_in_each_class, 3)/5\n","c3 = (0.4, 0.1, 0.7) + np.random.rand(samples_in_each_class, 3)/5\n","data = np.float32(np.concatenate((c1, c2, c3)))\n","colors = [\"red\"] * samples_in_each_class\n","colors.extend([\"green\"] * samples_in_each_class)\n","colors.extend([\"blue\"] * samples_in_each_class)\n","fig = plt.figure()\n","ax = Axes3D(fig)\n","ax.scatter(data[:, 0], data[:, 1], data[:, 2], c=colors)\n","labels = range(3*samples_in_each_class)"],"execution_count":5,"outputs":[{"output_type":"display_data","data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAb4AAAEuCAYAAADx63eqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n","text/plain":["<matplotlib.figure.Figure at 0x7fe4f786b908>"]},"metadata":{"tags":[]}}]},{"metadata":{"id":"Fb7XCR8ETL2w","colab_type":"text"},"cell_type":"markdown","source":["## Αρχικοποίηση και εκπαίδευση του SOM\n","Στη συνέχεια αρχικοποιούμε ένα χάρτη SOM με 10 x 10 νευρώνες και τον εκπαιδεύουμε με default παραμέτρους. Χρησιμοποιούμε τη magic `time` για να πάρουμε το χρόνο εκπαίδευσης."]},{"metadata":{"id":"1GQVZ626TL2z","colab_type":"code","colab":{}},"cell_type":"code","source":["n_rows, n_columns = 10, 10\n","som = somoclu.Somoclu(n_columns, n_rows, compactsupport=False)\n","%time som.train(data)"],"execution_count":0,"outputs":[]},{"metadata":{"id":"7tn6Ag8bTL28","colab_type":"text"},"cell_type":"markdown","source":["Δείτε [εδώ](https://somoclu.readthedocs.io/en/stable/reference.html) όλο το function reference της βιβλιοθήκης."]},{"metadata":{"id":"rkUcxjDWTL2-","colab_type":"text"},"cell_type":"markdown","source":["<table>\n","  <tr>\n","    <td bgcolor=\"#FCF8E3\"><font size=\"4\">Αν δεν σας τρέχει το train, δείτε το status των clouds γενικά και ως προς το Somoclu στο <a href=\"https://docs.google.com/document/d/1jL4gRag_LHbVCYIt5XVJ53iJPb6RZWi02rT5mPXiqEU/edit?usp=sharing\">FAQ</a> !!</font>\n","</td>\n","  </tr>\n","</table>"]},{"metadata":{"id":"t5Wpdj3rTL3E","colab_type":"text"},"cell_type":"markdown","source":["## Απεικόνιση SOM\n","\n","Η βιβλιοθήκη προσφέρει διάφορες δυνατότητες απεικόνισης του χάρτη."]},{"metadata":{"id":"7sLgAzlITL3I","colab_type":"text"},"cell_type":"markdown","source":["### Επίπεδα συνιστωσών\n","\n","H πρώτη είναι να δούμε τα επίπεδα συνιστωσών (component planes), ως προς τους (τρεις) άξονες των διαστάσεων εισόδου. To colorbar μας δείχνει ποιοι νευρώνες του codebook (δηλαδή των 10x10 νευρώνων εξόδου) είναι κοντά. Θυμόμαστε ότι οι νευρώνες εξόδου βρίσκονται σε χώρο ίδιων διαστάσεων με τα διανύσματα εισόδου. Τα 3 γραφήματα είναι για x, y και z."]},{"metadata":{"scrolled":true,"id":"1ZOV9jtNTL3K","colab_type":"code","colab":{}},"cell_type":"code","source":["som.view_component_planes(colorbar=True, bestmatches=True, figsize=(15, 15))"],"execution_count":0,"outputs":[]},{"metadata":{"id":"rjkhFJ36TL3T","colab_type":"text"},"cell_type":"markdown","source":["Η figsize ελέγχει τις διαστάσεις του γραφήματος (zoom), όχι τις διαστάσεις του grid.\n","\n","Η απεικόνιση με επίπεδα συνιστωσών είναι πολύ αναλυτική αλλά ίσως πολύ αναλυτική για να μας βοηθήσει στην ομαδοποίηση."]},{"metadata":{"id":"tBvWv2cTTL3V","colab_type":"text"},"cell_type":"markdown","source":["### U-matrix\n","\n","Στην πράξη χρησιμοποιούμε το U-matrix που μας δείχνει το πλέγμα του SOM και το χρωματίζει ανάλογα το πόσο απέχουν οι νευρώνες εξόδου μεταξύ τους (όπως είδαμε στο μάθημα και σύμφωνα με το colorbar)"]},{"metadata":{"id":"DJP9TBvQTL3a","colab_type":"code","colab":{}},"cell_type":"code","source":["som.view_umatrix(bestmatches=True, bestmatchcolors=colors, labels=labels, colorbar=True, figsize=(15, 15))"],"execution_count":0,"outputs":[]},{"metadata":{"id":"JoYt9XAlTL3e","colab_type":"text"},"cell_type":"markdown","source":["## Παρατηρήσεις για την απεικόνιση U-matrix\n","\n","- με `bestmatches=True` και `labels=labels` τυπωνουμε τα labes των δειγμάτων εισόδου που ανήκουν σε κάθε νευρώνα νικήτη (best matching unit - bmu). Προσοχή εδώ, στο local install του Somoclu ο χάρτης είναι διαδραστικός, στο notebook οχι. Προφανώς δεν ανήκει μόνο ένα δείγμα εισόδου σε κάθε bmu όπως φαίνεται εδώ αφου έχουμε 100 bmus και samples_in_each_class x 3 δείγματα εισόδου (πχ  600 για samples_in_each_class=200). Ο λόγος που εμφανίζεται έτσι είναι γιατί εμφανίζεται μόνο η ετικέτα του τελευταίου δείγματος που ανατίθεται στο κάθε bmu (φανταστείτε δηλαδή ότι κάτω από κάθε ετικέτα συνήθως υπάρχουν και άλλες).\n","- Οι κόκκινες περιοχές δείχνουν περιοχές με μεγάλες αποστάσεις και οι μπλε περιοχές δείχνουν περιοχές με μικρές αποστάσεις. Εποπτικά, βλέπουμε ότι έχει διαχωρίσει καλά σε τρία clusters τα δείγματα εισόδου. "]},{"metadata":{"id":"GuEK8n2LTL3f","colab_type":"text"},"cell_type":"markdown","source":["## Clustering των νευρώνων νικητών (bmus)\n","Η `bestmatchcolors=colors` βάζει χρώμα στα bmu που ανήκουν στο ίδιο cluster. Αυτό δεν γίνεται εποπτικά (\"με το μάτι\" από το U-matrix) αλλά ως εξής σε τρια βήματα:  \n","1. μέσω ανταγωνιστικής μάθησης έχουμε μάθει τα βάρη των νευρώνων νικητών (bmus). \n","2. εφαρμόζουμε έναν αλγόριθμο ομαδοποίησης όπως k-means για βρούμε ποιοι νευρώνες νικητές ανήκουν στο ίδιο cluster\n","3. χρωματίζουμε με το ίδιο χρώμα τους νευρώνες κάθε cluster\n","\n","Μπορούμε να ορίσουμε όποιον αλγόριθμο clustering θέλουμε για τα bmus του SOM:\n","```python\n","from sklearn.cluster import KMeans\n","algorithm = KMeans()\n","som.cluster(algorithm=algorithm)\n","```"]},{"metadata":{"id":"Bkm4EzLETL3i","colab_type":"text"},"cell_type":"markdown","source":["### Best matching units (BMUS)\n","\n","Μπορούμε να δούμε σε ποιο νευρώνα νικήτή ανήκει κάθε δείγμα εισόδου ως εξής:"]},{"metadata":{"id":"fYHb_UPRTL3l","colab_type":"code","colab":{}},"cell_type":"code","source":["bmus = som.bmus"],"execution_count":0,"outputs":[]},{"metadata":{"id":"i5z8K9QNTL3r","colab_type":"code","colab":{}},"cell_type":"code","source":["print(bmus.shape)"],"execution_count":0,"outputs":[]},{"metadata":{"id":"gqSxrfWCTL3z","colab_type":"text"},"cell_type":"markdown","source":["οι διαστάσεις είναι τα διανύσματα εισόδου επί τις διαστάσεις του grid μας (2 εφόσον είναι 10x10). Ο πίνακας bmus μας δείχνει τις συντεταγμένες στο grid του νευρώνα νικητή στον οποίο ανήκει το κάθε δείγμα εισόδου."]},{"metadata":{"id":"N-DlqmRtTL31","colab_type":"code","colab":{}},"cell_type":"code","source":["print(bmus)"],"execution_count":0,"outputs":[]},{"metadata":{"id":"ZY6Pgfb6TL35","colab_type":"text"},"cell_type":"markdown","source":["**Προσοχή**\n","- η σύμβαση του Somoclu είναι \\[κολόνα γραμμή\\] (ανάποδα από την Python)\n","- αν πάρετε τις μοναδικές εμφανίσεις των bmus μπορεί να είναι λιγότερες από 100 (10x10). Αυτό γιατί μπορεί να υπάρξουν νευρώνες εξόδου που δεν νίκησαν ποτέ και δεν τους έχουν ανατεθεί δείγματα εδώ.\n","\n","Μπορούμε να απαριθμήσουμε τα μοναδικά bmus και να πάρουμε άνα πίνακα που μας δείχνει σε ποιο αριθμό bmu ανήκει κάθε δείγμα εισόδου ως εξής:"]},{"metadata":{"id":"tGTe9wsZTL36","colab_type":"code","colab":{}},"cell_type":"code","source":["ubmus, indices = np.unique(bmus, return_inverse=True, axis=0)\n","# το return_inverse επιστρέφει και τους δείκτες indices"],"execution_count":0,"outputs":[]},{"metadata":{"id":"liqFO73_TL4E","colab_type":"text"},"cell_type":"markdown","source":["Τυπώνουμε τα μοναδικά bmus. Το \\[0 0\\] είναι το bmu 0, το \\[0 1\\] είναι το bmu 1, το \\[0 2\\] είναι το bmu 2  κοκ. "]},{"metadata":{"id":"ey7qHCZzTL4F","colab_type":"code","colab":{}},"cell_type":"code","source":["print(ubmus)"],"execution_count":0,"outputs":[]},{"metadata":{"id":"wCijrdW-TL4O","colab_type":"text"},"cell_type":"markdown","source":["Τυπώνουμε τέλος για κάθε δείγμα εισόδου σε ποιο αριθμό bmu ανήκει. (πιο βολικό από τις συντεταγμένες των bmus)"]},{"metadata":{"scrolled":true,"id":"EGRVcxOSTL4R","colab_type":"code","colab":{}},"cell_type":"code","source":["print(indices)"],"execution_count":0,"outputs":[]}]}