eeg_microstates3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Information-theoretic analysis of EEG microstates in Python\n",
    "\n",
    "Our software is provided as a single Python source code file (eeg_microstates.py) that contains all functions necessary to:  \n",
    "* import EEG data in the '.edf' file format\n",
    "* pre-process (band-pass filter) the EEG data\n",
    "* compute microstate maps by the modified K-means algorithm [1]\n",
    "* compute microstate time series by competitively back-fitting the microstate maps into the EEG data\n",
    "* perform information-theoretical calculations we recently introduced [2]\n",
    "\n",
    "[1] Pascual-Marqui RD, Michel CM, Lehmann D. Segmentation of brain electrical activity into microstates: model estimation and validation. IEEE Trans Biomed Eng 42 (1995) 658–665.\n",
    "\n",
    "[2] von Wegner F, Tagliazucchi E, Laufs H. Information-theoretical analysis of resting state EEG microstate sequences - non-Markovianity, non-stationarity and periodicities. NeuroImage 158 (2017) 99–111.\n",
    "\n",
    "[3] von Wegner F, Laufs H. Information-theoretical analysis of EEG microstate sequences in Python. Front Neuroinformatics (2018) doi: 10.3389/fninf.2018.00030.\n",
    "\n",
    "[4] von Wegner F, Knaut P, Laufs H. EEG microstate sequences from different clustering algorithms are information-theoretically invariant. Front Comp Neurosci (2018) doi: 10.3389/fncom.2018.00070.\n",
    "\n",
    "\n",
    "### Github link of this project\n",
    "\n",
    "The current version of the software can be found on the [GitHub project page](https://github.com/Frederic-vW/eeg_microstates/).\n",
    "\n",
    "### Current literature about EEG microstates and information theory\n",
    "\n",
    "To find out about current publications and directions in microstate research, in particular related to information-theoretical approaches, we provide an active link to the PubMed page:  \n",
    "https://www.ncbi.nlm.nih.gov/pubmed/?term=eeg+microstates+information+theory\n",
    "\n",
    "## Background\n",
    "\n",
    "Electroencephalography (EEG) records the brain's electrical potential at discrete spatial sampling sites, defined by the locations of the electrodes on the skull. We also call the potential distribution at a given time a topography, or a map. The idea behind the microstate algorithm is to represent the EEG time series by a small number of maps that yield an optimum representation of the full data set. As in other areas of data compression, optimality can be defined in different ways. For EEG microstates, the criterion of choice is that of maximum explained variance. To this end, a (large) set of EEG samples is fed into a clustering algorithm whose output is a reduced set of representative EEG topographies, also called termed microstates, or microstate maps.  \n",
    "The EEG time series can then be represented by a sequence of microstate labels, where each time point of the original EEG data is mapped to the microstate with the maximum spatial correlation to the actual EEG topography. Polarity is usually ignored, so the squared correlation coefficient between each candidate microstate map and the EEG topography is used as a similarity measure.  \n",
    "With a set of representative microstates at hand, the multi-channel EEG data set can be compressed into a time series of symbols, i.e. microstate labels. Most frequently, four microstates are used to compress the EEG data set, and the microstates are labelled A, B, C, D. The EEG data set is therefore transformed into a microstate sequence that could read like ...AAABBDCDABBDAB...  \n",
    "Due to the special, non-metric nature of the new state space {A, B, C, D}, many common time-series methods such as the autocorrelation function cannot be applied to the microstate time series. We therefore test several entropy-based measures based on Shannon's information theory to quantify these time series.  \n",
    "\n",
    "The details of the procedure are given in the following sections and the whole procedure is illustrated in the figure below.  \n",
    "Step 1: The EEG data set is submitted to the modified K-means algorithm to obtain the microstate maps A-D. The figure below shows a section of resting state EEG for a selection of the 30 channels recorded. At each time point, the spatial standard deviation of the EEG topography is also called global field power (GFP, blue). Topographies at local maxima of the GFP time series (MAX, red dots) define the input to the modified K-means clustering algorithm.  \n",
    "Step 2: The microstate time series is computed by competitive back-fitting of the microstate maps into the EEG time series.  \n",
    "Step 3: Several information-theoretical tests are run on the microstate sequence and finally, the time-lagged mutual information function and a first-order Markov confidence interval are computed and displayed.  \n",
    "\n",
    "<img src=\"Figure_1.jpg\" width=800>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import package\n",
    "\n",
    "In order to work with the package, the following import statement can be used in Python standalone scripts, e.g. for batch processing, or in any Python-based EEG software to add extra functionality:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from eeg_microstates3 import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (1) EEG data input"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The package contains a basic [EDF](http://www.edfplus.info/specs/edf.html) file format reader. EEG data is returned as a NumPy array with time running along the rows (first index) and channels across the columns (2nd index), i.e. the array shape is (n_samples, n_channels). If our code is interfaced with other Python programs, EEG data must be provided in that shape in order to be processed correctly by our package functions.  \n",
    "The return variables of the function read_edf are (1) a list of channel names, (2) the sampling rate in Hz, and (3) the EEG data in the mentioned format.  \n",
    "To run this tutorial, you must have the test.edf EEG file in your current folder. The file is part of the GitHub repository. The record contains 192 sec. from an eyes-closed resting state experiment of a healthy male subject recorded with a 30 channel EEG cap using a standard 10-10 electrode configuration. The sampling rate is 250 Hz and the data is band-pass filtered to the 1-40 Hz range."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "File: test.edf\n",
      "Sampling rate: 250.00 Hz\n",
      "Data shape: 48000 samples x 30 channels\n",
      "Channels:\n",
      "Fp1, Fp2, F3, F4, C3, C4, P3, P4, O1, O2, F7, F8, T7, T8, P7, P8, Fz, Cz, Pz, AFz, AF3, AF4, FC3, FC4, FT9, FT10, TP9, TP10, CP5, CP6, \n"
     ]
    }
   ],
   "source": [
    "filename = \"test.edf\"\n",
    "chs, fs, data_raw = read_edf(filename)\n",
    "print(\"File: {}\".format(filename))\n",
    "print(\"Sampling rate: {:.2f} Hz\".format(fs))\n",
    "print(\"Data shape: {:d} samples x {:d} channels\".format(data_raw.shape[0], data_raw.shape[1]))\n",
    "print(\"Channels:\")\n",
    "for ch in chs:\n",
    "    print(f\"{ch.decode():s}\", end=', ')\n",
    "print(\"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (2) Pre-processing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "EEG data is usually band-pass filtered before microstate clustering. The package contains a zero-phase 6-th order Butterworth band-pass filter as provided by SciPy. The input is (1) the data to be filtered as a NumPy array of shape (n_samples, n_channels), (2) a tuple defining the desired frequency band in Hz, and (3) the EEG sampling rate in Hz."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(48000, 30)\n"
     ]
    }
   ],
   "source": [
    "data = bp_filter(data_raw, f_lo=1, f_hi=35, fs=fs)\n",
    "print(data.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (3) Visualize data (PCA-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A basic data visualization tool is provided to get an idea of the temporal structure of the EEG data set. We use principle component analysis as provided in the scikit-learn package and plot the time course of the first principal component to represent the dominant component of the data set. Using the test.edf file, we find amplitude modulated oscillations, mainly in the alpha frequency band (8-12), as expected in a wakeful resting condition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pca = PCA(copy=True, n_components=1, whiten=False)\n",
    "pca1 = pca.fit_transform(data)[:,0]\n",
    "plot_data(pca1, fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Alternatively, only a segment of data can be visualized to better recognize alpha oscillations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "t0, t1 = 1000, 3000\n",
    "plot_data(pca1[t0:t1], fs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (4) Clustering of the microstate maps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are now ready to apply the microstate clustering algorithm to the filtered data. We implemented five different clustering methods (AAHC, modified k-means, k-medoids, PCA, Fast-ICA), as detailed in [4].  \n",
    "Here, we will run the modified K-means algorithm as given by Pascual-Marqui [1], implemented as the function 'kmeans' in the eeg_microstates.py file [3]. The function variables should be self-explaining and are further documented in the source file following the PEP 8 styleguide.    \n",
    "As the initial microstate ordering returned by the stochastic K-means algorithm is random, we provide a tool to re-label the microstates interactively. The standard ordering follows the geometry of the maps. Map A has a diagonal direction connecting the upper (frontal) left and lower (occipital) right corners, map B is diagonal in the opposite direction, map C has a horizontal orientation and map D is often circular. Sometimes, slightly different maps are generated. We mostly observed the occurrence of a map D with a vertical axis instead of a circular pattern. In this case, you can either accept the results and proceed, or re-cluster the data set as the results of the K-means algorithm differ between runs.\n",
    "Modified K-means clustering and all information-theoretical computations can be run for any number of microstates. The desired number of microstates is given by the user as the argument 'n_maps' of the 'kmeans' function. If n_maps is not provided, the default value n_maps=4 is used."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Clustering algorithm: kmeans\n"
     ]
    }
   ],
   "source": [
    "mode = [\"aahc\", \"kmeans\", \"kmedoids\", \"pca\", \"ica\"][1]\n",
    "print(f\"Clustering algorithm: {mode:s}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[+] Data format for clustering [GFP peaks, channels]: 4130 x 30\n",
      "\n",
      "[+] Clustering algorithm: mod. K-MEANS.\n",
      "\tK-means run 1/5 converged after 33 iterations.\n",
      "\tK-means run 2/5 converged after 39 iterations.\n",
      "\tK-means run 3/5 converged after 43 iterations.\n",
      "\tK-means run 4/5 converged after 31 iterations.\n",
      "\tK-means run 5/5 converged after 94 iterations.\n",
      "[+] Computation time: 1.41 sec\n",
      "\n",
      "[+] Microstate back-fitting:\n",
      "data_norm:  (48000, 30)\n",
      "data_cluster_norm:  (4130, 30)\n",
      "maps:  (4, 30)\n",
      "\n",
      "[+] Global explained variance GEV = 0.686\n",
      "GEV_0: 0.250\n",
      "GEV_1: 0.174\n",
      "GEV_2: 0.102\n",
      "GEV_3: 0.160\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Re-ordered labels: 0, 1, 2, 3\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_maps = 4\n",
    "locs = []\n",
    "maps, x, gfp_peaks, gev = clustering(data, fs, chs, locs, mode, n_maps, interpol=False, doplot=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also inspect the source code of the function:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import inspect\n",
    "#c=inspect.getsourcelines(kmeans) # activate by uncomment\n",
    "#for l in c[0]: print(l)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When the algorithm terminates, the global explained variance (GEV) of the computed microstate time series is shown. For the test file, the GEV should give approximately 0.73. In other words, the microstate sequence explains ca. 73% of the electrical potential's spatial variance over time.  \n",
    "\n",
    "The first return variable of the kmeans function is an array containing the microstate maps row-wise. The second return variable is the microstate sequence corresponding to the EEG data that was used as input. The sequence was generated as part of the kmeans optimizations by back-fitting into the EEG time series, and the sequence corresponding to the best K-means run is returned.  \n",
    "On this sequence all subsequent tests will be run. The sequence contains the microstate labels as numbers, $\\{A, B, C, D\\} \\rightarrow \\{0, 1, 2, 3\\}$, in order to accelerate information-theoretical computations. Using numbers, the microstate label can directly be used as an index in arrays and matrices. If you want to test microstate sequences generated by other another software, the sequence must be converted to numerical values $0, ..., n_{maps}-1$ before using the functions of this package."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (5) Basic microstate statistics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now look at some basic statistical properties of the microstate sequence returned by the kmeans function. The empirical symbol distribution $\\hat{p}$ quantifies the relative frequency of each microstate map. It is equivalent to the so-called 'ratio of time covered' (RTT) used in some EEG microstate publications. The transition matrix $\\hat{T}$ contains the transition probabilities between all pairs of microstate labels. To get a feeling of the temporal dynamics of the EEG's spatial variance, we also compute the number of GFP peaks per second. The value is roughly 2 x the dominant EEG frequency band. In our case, the dominant frequency is a 10 Hz alpha rhythm and we observe approx. 21.5 GFP peaks per second."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Empirical symbol distribution (RTT):\n",
      "\n",
      "p_0 = 0.237\n",
      "p_1 = 0.282\n",
      "p_2 = 0.219\n",
      "p_3 = 0.261\n",
      "\n",
      "Empirical transition matrix:\n",
      "\n",
      "|0.7770.0800.0880.055|\n",
      "|0.0630.8030.0580.076|\n",
      "|0.0600.0810.7610.098|\n",
      "|0.0850.0720.0570.786|\n",
      "\n",
      "GFP peaks per sec.: 21.51\n",
      "\n",
      "Global explained variance (GEV) per map:\n",
      "GEV(ms-0) = 0.25\n",
      "GEV(ms-1) = 0.17\n",
      "GEV(ms-2) = 0.10\n",
      "GEV(ms-3) = 0.16\n",
      "\n",
      "total GEV: 0.686\n"
     ]
    }
   ],
   "source": [
    "p_hat = p_empirical(x, n_maps)\n",
    "T_hat = T_empirical(x, n_maps)\n",
    "print(\"\\nEmpirical symbol distribution (RTT):\\n\")\n",
    "for i in range(n_maps):\n",
    "    print(f\"p_{i:d} = {p_hat[i]:.3f}\")\n",
    "print(\"\\nEmpirical transition matrix:\\n\")\n",
    "print_matrix(T_hat)\n",
    "pps = len(gfp_peaks) / (len(x)/fs)  # peaks per second\n",
    "print(f\"\\nGFP peaks per sec.: {pps:.2f}\")\n",
    "print(\"\\nGlobal explained variance (GEV) per map:\")\n",
    "for i, g in enumerate(gev): print(f\"GEV(ms-{i:d}) = {gev[i]:.2f}\")\n",
    "print(f\"\\ntotal GEV: {gev.sum():.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (6) Shannon entropy and entropy rate of the microstate sequence"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The information-theoretical characterization of the microstate time series starts with the calculation of the empirical Shannon entropy $\\hat{H}$ of the sequence:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$$\n",
    "\\begin{align}\n",
    "\\hat{H} & = -\\sum_i \\hat{p}(i) \\log \\hat{p}(i)\n",
    "\\end{align}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "where the index $i$ runs over the microstate labels. The base of the logarithm is not important. We use the natural logarithm to the base $e$ (Euler's number) and therefore measure information in 'nats', in analogy to bits when using $\\log_2$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Empirical entropy H = 1.38 (max. entropy: 1.39)\n"
     ]
    }
   ],
   "source": [
    "h_hat = H_1(x, n_maps)\n",
    "h_max = max_entropy(n_maps)\n",
    "print(f\"\\nEmpirical entropy H = {h_hat:.2f} (max. entropy: {h_max:.2f})\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The maximum entropy possible for any sequence composed of four symbols is $\\log(4) \\approx 1.39$, and the minimum entropy is zero. We see that the test sequence almost achieves the maximum entropy. As entropy can be interpreted as the amount of surprise or randomness, this result suggests a lack of predictable temporal features in the microstate sequence. However, we will show that the sequence does contain temporal structure, mainly in the form of periodicities in the range of the main EEG frequency bands."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, the entropy rate of the sequence is estimated from a set of joint Shannon entropies $H_k=H(\\mathbf{X}_{n}^{(k)})$ for different history lengths $k$. The estimated entropy rate is the slope of the linear fit $k$ vs. $H_k$. The intercept is the so-called excess entropy, which is not further investigated here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Empirical entropy rate h = 0.73\n",
      "Theoretical MC entropy rate h = 0.76\n"
     ]
    }
   ],
   "source": [
    "h_rate, _ = excess_entropy_rate(x, n_maps, kmax=8, doplot=True)\n",
    "h_mc = mc_entropy_rate(p_hat, T_hat)\n",
    "print(f\"\\nEmpirical entropy rate h = {h_rate:.2f}\")\n",
    "print(f\"Theoretical MC entropy rate h = {h_mc:.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (7) Markov tests"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will first test the Markov properties of order 0, 1 and 2 for the EEG test file. For the details of the testing procedure, see the attached manuscript and the code documentation in the source file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\ZERO-ORDER MARKOVIANITY:\n",
      "p: 0.00e+00 | t: 59954.303 | df: 9.0\n",
      "\n",
      "FIRST-ORDER MARKOVIANITY:\n",
      "p: 3.83e-123 | t: 695.807 | df: 36.0\n",
      "\n",
      "SECOND-ORDER MARKOVIANITY:\n",
      "p: 2.38e-81 | t: 742.497 | df: 144.0\n",
      "\n",
      "GEOMETRIC DISTRIBUTION of lifetimes:\n",
      "\n",
      "\n",
      "Testing the distribution of symbol # 0\n",
      "p: 1.21e-16 | t: 136.814 | df: 27.0\n",
      "G-test (log-likelihood) p: 2.31e-06, g: 74.737, df: 27.0\n",
      "G-test (Pearson Chi2) p: 2.38e-05, g: 67.687, df: 27.0\n",
      "\n",
      "Testing the distribution of symbol # 1\n",
      "p: 1.39e-65 | t: 390.859 | df: 28.0\n",
      "G-test (log-likelihood) p: 8.90e-31, g: 215.175, df: 28.0\n",
      "G-test (Pearson Chi2) p: 8.43e-27, g: 194.229, df: 28.0\n",
      "\n",
      "Testing the distribution of symbol # 2\n",
      "p: 2.91e-13 | t: 123.242 | df: 30.0\n",
      "G-test (log-likelihood) p: 5.48e-05, g: 69.609, df: 30.0\n",
      "G-test (Pearson Chi2) p: 5.12e-04, g: 62.079, df: 30.0\n",
      "\n",
      "Testing the distribution of symbol # 3\n",
      "p: 2.39e-23 | t: 180.145 | df: 30.0\n",
      "G-test (log-likelihood) p: 5.98e-09, g: 96.762, df: 30.0\n",
      "G-test (Pearson Chi2) p: 6.35e-08, g: 90.097, df: 30.0\n"
     ]
    }
   ],
   "source": [
    "alpha = 0.01\n",
    "p_M0 = testMarkov0(x, n_maps, alpha)\n",
    "p_M1 = testMarkov1(x, n_maps, alpha)\n",
    "p_M2 = testMarkov2(x, n_maps, alpha)\n",
    "p_vals_geo = geoTest(x, n_maps, 1000./fs, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The statistical Markovianity test show that our test sequence does not have the Markov property of the tested orders. This means that there are memory effects in the series which we will further explore."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (8) Stationarity test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we will test if the transition matrix $\\hat{T}$ is time-stationary or if the matrix elements change significantly over time. You can provide a specific window length L for testing or use the default value of 5000 samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter block size: 2500\n",
      "\n",
      "CONDITIONAL HOMOGENEITY (three-way table):\n",
      "Split data in r = 19 blocks of length 2500.\n",
      "p: 2.10e-12 | t: 392.765 | df: 216.0\n"
     ]
    }
   ],
   "source": [
    "try:\n",
    "    L = int(input(\"Enter block size: \"))\n",
    "except:\n",
    "    L = 5000\n",
    "p3 = conditionalHomogeneityTest(x, n_maps, L, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We find that the given transition matrix is not time-stationary."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (9) Symmetry test"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will now test if the transition matrix is symmetric. In other words, we test if the transitions $A \\rightarrow B$ and $B \\rightarrow A$ occur with the same probability or not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "SYMMETRY:\n",
      "p: 2.71e-55 | t: 269.547 | df: 6.0\n"
     ]
    }
   ],
   "source": [
    "p4 = symmetryTest(x, n_maps, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The transition matrix of the test microstate sequence is not symmetric."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (10) Markov surrogate example"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To test our statistical procedures and to construct surrogate data sets from which confidence intervals for further information-theoretical quantities can be computed, we will synthesize first-order Markov chains with symbol distribution and transition matrix identical to the experimental EEG data set. This is possible because a first-order Markov chain is completely defined by an initial symbol distribution and a transition matrix for first-order (single time step) transitions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Surrogate symbol distribution:\n",
      "\n",
      "p_0 = 0.228\n",
      "p_1 = 0.283\n",
      "p_2 = 0.220\n",
      "p_3 = 0.268\n",
      "\n",
      "Surrogate transition matrix:\n",
      "\n",
      "|0.7690.0830.0920.057|\n",
      "|0.0640.8030.0590.075|\n",
      "|0.0570.0790.7650.100|\n",
      "|0.0830.0730.0530.791|\n"
     ]
    }
   ],
   "source": [
    "x_mc = surrogate_mc(p_hat, T_hat, n_maps, len(x))\n",
    "p_surr = p_empirical(x_mc, n_maps)\n",
    "T_surr = T_empirical(x_mc, n_maps)\n",
    "print(\"\\nSurrogate symbol distribution:\\n\")\n",
    "for i in range(n_maps):\n",
    "    print(f\"p_{i:d} = {p_surr[i]:.3f}\")\n",
    "print( \"\\nSurrogate transition matrix:\\n\" )\n",
    "print_matrix(T_surr)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We observe that the symbol distribution and the transition matrix are equivalent, within some stochastic margin, to the experimental data. A quantitative analysis of Markov surrogate errors is contained in [2]."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (11) Markov tests for surrogate Markov data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "It will now be shown that the Markov surrogates fulfill the expected properties, or vice versa, that our Markovianity and stationarity tests work well with our surrogate data. This means that we can analyze experimental data series to find memory effects beyond a first-order Markov model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\ZERO-ORDER MARKOVIANITY:\n",
      "p: 0.00e+00 | t: 59982.540 | df: 9.0\n",
      "\n",
      "FIRST-ORDER MARKOVIANITY:\n",
      "p: 8.83e-01 | t: 26.268 | df: 36.0\n",
      "\n",
      "SECOND-ORDER MARKOVIANITY:\n",
      "p: 3.53e-01 | t: 149.818 | df: 144.0\n",
      "\n",
      "GEOMETRIC DISTRIBUTION of lifetimes:\n",
      "\n",
      "\n",
      "Testing the distribution of symbol # 0\n",
      "p: 4.78e-01 | t: 28.756 | df: 29.0\n",
      "G-test (log-likelihood) p: 9.91e-01, g: 14.005, df: 29.0\n",
      "G-test (Pearson Chi2) p: 9.95e-01, g: 13.251, df: 29.0\n",
      "\n",
      "Testing the distribution of symbol # 1\n",
      "p: 1.61e-02 | t: 50.167 | df: 31.0\n",
      "G-test (log-likelihood) p: 8.49e-01, g: 23.011, df: 31.0\n",
      "G-test (Pearson Chi2) p: 8.78e-01, g: 22.153, df: 31.0\n",
      "\n",
      "Testing the distribution of symbol # 2\n",
      "p: 7.03e-02 | t: 48.015 | df: 35.0\n",
      "G-test (log-likelihood) p: 9.80e-01, g: 20.073, df: 35.0\n",
      "G-test (Pearson Chi2) p: 9.94e-01, g: 17.497, df: 35.0\n",
      "\n",
      "Testing the distribution of symbol # 3\n",
      "p: 8.09e-01 | t: 26.701 | df: 34.0\n",
      "G-test (log-likelihood) p: 1.00e+00, g: 13.038, df: 34.0\n",
      "G-test (Pearson Chi2) p: 1.00e+00, g: 12.138, df: 34.0\n"
     ]
    }
   ],
   "source": [
    "p_M0_mc = testMarkov0(x_mc, n_maps, alpha)\n",
    "p_M1_mc = testMarkov1(x_mc, n_maps, alpha)\n",
    "p_M2_mc = testMarkov2(x_mc, n_maps, alpha)\n",
    "p_vals_geo_mc = geoTest(x_mc, n_maps, 1000./fs, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Markov surrogate clearly fulfills the expected theoretical properties. The zero-order Markov hypothesis is rejected while the first-order and second-order properties are accepted. The theoretically equivalent, but numerically different test for geometric lifetime distributions asserts the geometric lifetime distribution of all microstate maps."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (12) Stationarity test for surrogate Markov data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We expect the Markov surrogate to be time-stationary as the same transition matrix has been used to generate the entire sequence. The test results show that indeed, the Markov surrogate is correctly classified as time-stationary. This means that we can identify non-stationarity in experimental data sets statistically."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter block size: 3000\n",
      "\n",
      "CONDITIONAL HOMOGENEITY (three-way table):\n",
      "Split data in r = 16 blocks of length 3000.\n",
      "p: 8.58e-01 | t: 159.814 | df: 180.0\n"
     ]
    }
   ],
   "source": [
    "try:\n",
    "    L = int(input(\"Enter block size: \"))\n",
    "except:\n",
    "    L = 5000\n",
    "p3_mc = conditionalHomogeneityTest(x_mc, n_maps, L, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (13) Symmetry test for surrogate Markov data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the experimental EEG transtion matrix is used to generate the Markov surrogate, the surrogate's transition matrix is also asymmetric."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "SYMMETRY:\n",
      "p: 1.03e-63 | t: 308.866 | df: 6.0\n"
     ]
    }
   ],
   "source": [
    "p4_mc = symmetryTest(x_mc, n_maps, alpha)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (14) Time-lagged mutual information: empirical vs. Markov"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Mutual information for time lag k, $I(k)$, is defined as the difference between the entropy (uncertainty) of the microstate sequence at time point $t+k$, and the conditional entropy at time point $t+k$ given the state at time $t$. Mutual information thus quantifies the uncertainty about the state at $t+k$ that is not represented by the previous time point $t$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$$\n",
    "\\begin{align}\n",
    "I(k) & = H(X_{t+k}) - H(X_{t+k} \\mid X_k)\n",
    "\\end{align}\n",
    "$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As an analogy to the autocorrelation function (ACF) for metric time series, we will call this information-theoretic quantity the autoinformation function (AIF). Whilte the ACF uses linear correlation as a dependency measure for metric time series, the AIF uses mutual information to quantify dependencies in symbolic sequences.  \n",
    "We can now compute the time-lagged mutual information function of the microstate sequence and compare it to a set of Markov surrogates, from which a confidence interval under the first-order Markovian null hypothesis is constructed. We here use 10 Markov surrogates and compute the two-sided 99% percentile.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mutual information lag: 100\n",
      "\n",
      "Computing n = {n_mc:d} Markov surrogates...\n",
      "\n",
      "\n",
      "surrogate MC # 1/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 2/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 3/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 4/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 5/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 6/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 7/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 8/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 9/10\n",
      "mutual information lag: 100\n",
      "\n",
      "surrogate MC # 10/10\n",
      "mutual information lag: 100\n"
     ]
    }
   ],
   "source": [
    "l_max = 100\n",
    "n_mc = 10\n",
    "aif = mutinf(x, n_maps, l_max)\n",
    "print(f\"\\nComputing n = {n_mc:d} Markov surrogates...\\n\")\n",
    "aif_array = mutinf_CI(p_hat, T_hat, len(x), alpha, n_mc, l_max)\n",
    "pct = np.percentile(aif_array,[100.*alpha/2.,100.*(1-alpha/2.)],axis=0)\n",
    "print(\"\\nMarkov surrogates and AIF confidence interval computed\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.ioff()\n",
    "fig = plt.figure(1, figsize=(20,5))\n",
    "fig.patch.set_facecolor('white')\n",
    "t = np.arange(l_max)*1000./fs\n",
    "plt.semilogy(t, aif, '-sk', linewidth=3, label='AIF (EEG)')\n",
    "plt.semilogy(t, pct[0,:], '-k', linewidth=1)\n",
    "plt.semilogy(t, pct[1,:], '-k', linewidth=1)\n",
    "plt.fill_between(t, pct[0,:], pct[1,:], facecolor='gray', alpha=0.5, label='AIF (Markov)')\n",
    "plt.xlabel(\"time lag [ms]\", fontsize=20, fontweight=\"bold\")\n",
    "plt.ylabel(\"AIF [nats]\", fontsize=20, fontweight=\"bold\")\n",
    "plt.legend(fontsize=22)\n",
    "ax = plt.gca()\n",
    "ax.tick_params(axis='both', labelsize=18)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The figure generated by the code above shows the autoinformation function (AIF) along with Markovian confidence intervals. The AIF summarizes memory effects in the test microstate sequence as identified by mutual information. It is observed that the AIF of experimental data (black line and squares) shows clear periodicities and comparison with the Markovian confidence interval (gray-shaded area) shows that these peaks are statistically significant.  \n",
    "We can thus show that EEG microstates recur in a regular way. In other words, the microstate sequences passes through the graph of state transitions in a circular way. The transitions do not do so in a completely deterministic way, as any state transition is possible at any time, but we show that there are preferred closed paths along the state transition graph."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## (15) Summary and reporting results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The results can be summarized for scientific reporting as follows. Please note that *fixed numerical values are shown here*, i.e. the results of a specifc run of the algorithm. As clustering results vary slightly between runs, the exact numerical values may differ from the results obtained above.\n",
    "\n",
    "1. We analyzed 192 seconds of resting state EEG sampled at 250 Hz. The 1-35 Hz band-pass filtered data set showed  21.51 GFP peaks per second and was clustered into 4 microstates, labelled A-D.\n",
    "\n",
    "2. Static properties:\n",
    "    * The microstate distribution was $p(A)=0.222$, $p(B)=0.240$, $p(C)=0.264$, and $p(D)=0.275$.\n",
    "    * The four microstates explained 69% of the total data variance (GEV).\n",
    "    * In information-theoretical terms, the four microstates produced a Shannon entropy of $H=1.38$ nats, where the upper bound for a 4-state process is $H_{max}=1.39$ nats. The empirical entropy rate was $h=0.73$ nats, and the expected value for the analogous first-order Markov process is $h=0.76$ nats.\n",
    "\n",
    "3. Dynamic properties:\n",
    "    * The transition probabilities between all pairs of microstates are summarized in the transition matrix T:\n",
    "    \\begin{equation*}\n",
    "    T =  \\begin{vmatrix}\n",
    "    0.771 & 0.094 & 0.058 & 0.076 \\\\\n",
    "\t0.063 & 0.769 & 0.112 & 0.056 \\\\\n",
    "\t0.063 & 0.055 & 0.786 & 0.096 \\\\\n",
    "\t0.070 & 0.072 & 0.061 & 0.797 \\\\\n",
    "    \\end{vmatrix}\n",
    "    \\end{equation*}\n",
    "    * The transition matrix T proved to be asymmetric ($p \\lt 10^{-80}$).\n",
    "\n",
    "4. Information-theoretical properties. Using a significance level of $\\alpha = 0.01$, we found that:\n",
    "    * The microstate sequence does not fulfill the Markov property of order 0, 1, or 2. We conclude that the microstate sequence has extended memory effects.\n",
    "    * During the EEG recording, the transition matrix of the microstate sequence is not constant (it is non-stationary), when tested for a partition into 9 blocks and 5000 samples per block ($p=3.37 \\times 10^{-6}$). This result should be taken as a precaution when using methods that require time stationarity (e.g. long-range memory interpretations of Hurst exponents).\n",
    "    * Non-Markovianity and non-stationarity are at least partially explained by a periodic recurrence of the microstates, as shown by the time-lagged mutual information function (or autoinformation function, AIF). For time lags up to 400 msec., the EEG's AIF lies above the (first-order) Markov confidence interval and shows distinct peaks at multiples of approx. 50 msec (half the period of the underlying alpha rhythm).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Code acceleration"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you ran the code above on your computer, you will have noticed that the AIF calculation determines the computation time of the procedure. The code can be accelerated using Cython, Numba or an external Julia function, for instance. To maximize portability and minimize the number of operating system dependencies, we did not include these accelerations in the code. However, the mentioned methods are implemented with little effort. Accelerated code can also be obtained from the author.\n",
    "The eeg_microstates.py script can be run from the console using the -mc parameter that defines the number of Markov surrogates. The number of surrogates will define the width and smoothness of the confidence interval. Empirically, we suggest a minimum number of 50-100 surrogates."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have arrived here, you should have completed the tutorial successfully. You should be familiar with the functionality provided by this software and have a good impression how to process larger data sets to obtain group statistics."
   ]
  }
 ],
 "metadata": {
  "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": "3.6.9"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}