[enh] update tutorial with more links, pictures

Author: emdupre

MAIN_tutorial_intro_to_nilearn.ipynb

@@ -43,10 +43,10 @@
     "import os\n",
     "import pandas as pd\n",
     "\n",
-    "data_dir = '/home/Desktop/MAIN_tutorial/'\n",
+    "data_dir = '/home/emdupre/Desktop/MAIN_tutorial/'\n",
     "participants = 'participants.tsv'\n",
     "phenotypic_data = pd.read_csv(os.path.join(data_dir, participants), sep='\\t')\n",
-    "phenotypic.head()"
+    "phenotypic_data.head()"
    ]
   },
   {
@@ -229,6 +229,41 @@
     "print(fmri_masked.shape)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "One way to think about what just happened is to look at it visually:\n",
+    "\n",
+    "![](http://nilearn.github.io/_images/masking.jpg)\n",
+    "\n",
+    "There are many other strategies in Nilearn [for masking data and for generating masks](http://nilearn.github.io/manipulating_images/manipulating_images.html#computing-and-applying-spatial-masks). I'd encourage you to spend some time exploring the documentation for these !"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "We can also [display this time series](http://nilearn.github.io/auto_examples/03_connectivity/plot_adhd_spheres.html#display-time-series) to get an intuition of how the whole brain signal is changing over time.\n",
+    "\n",
+    "We'll display the first three voxels by slicing the matrix using the colon-notation. You can also find more information on [how to slice arrays here](https://docs.scipy.org/doc/numpy-1.13.0/reference/arrays.indexing.html#basic-slicing-and-indexing)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "plt.plot(fmri_masked[:, :3])\n",
+    "\n",
+    "plt.title('Voxel Time Series')\n",
+    "plt.xlabel('Scan number')\n",
+    "plt.ylabel('Normalized signal')\n",
+    "plt.tight_layout()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -281,6 +316,17 @@
     "print(fmri_matrix.shape)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### [Compute and display a correlation matrix](http://nilearn.github.io/auto_examples/03_connectivity/plot_signal_extraction.html#compute-and-display-a-correlation-matrix)\n",
+    "\n",
+    "Now that we have a matrix, we'd like to create a _connectome_. A connectome is a map of the connections in the brain. Since we're working with functional data, however, we don't have access to actual connections. Instead, we'll use a measure of statistical dependency to infer the (possible) presence of a connection.\n",
+    "\n",
+    "Here, we'll use Pearson's correlation as our statistical dependency and compare how all of our ROIs from our chosen parcellation relate to one another."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -292,6 +338,30 @@
     "correlation_measure"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "correlation_matrix = correlation_measure.fit_transform([fmri_matrix])\n",
+    "correlation_matrix"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "\n",
+    "correlation_matrix = correlation_matrix[0]\n",
+    "# Mask the main diagonal for visualization:\n",
+    "# np.fill_diagonal(correlation_matrix, 0)\n",
+    "plotting.plot_matrix(correlation_matrix)"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,