Merge pull request #146 from alan-turing-institute/module03-updates

Updates to module 03

module03_research_data_in_python/03_03_plotting_with_matplotlib.ipynb

@@ -926,7 +926,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "We can plot all the data in the dataframe seperately, but that isn't always useful!"
+    "We can plot all the data in the dataframe separately, but that isn't always useful!"
    ]
   },
   {
@@ -948,21 +948,39 @@
     }
    ],
    "source": [
-    "df.plot(subplots=True);"
+    "df.plot(subplots=True)"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "And can produce more meaningful and useful visualisations either with our generated lists or by accessing the dataframe directly"
+    "Let's produce some more meaningful and useful visualisations by accessing the dataframe directly.\n",
+    "\n",
+    "We start by discarding any rows with an invalid (negative) standard deviation."
    ]
   },
   {
    "cell_type": "code",
    "execution_count": 32,
    "metadata": {},
    "outputs": [],
+   "source": [
+    "df = df[df[\"deviation\"] > 0]"
+   ]
+  },
+  {
+    "cell_type": "markdown",
+    "metadata": {},
+    "source": [
+     "Next we use the dataframe to construct some useful lists."
+    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 33,
+   "metadata": {},
+   "outputs": [],
    "source": [
     "deviation = df[\"deviation\"].tolist()  # Get the dataframe column (series) as a list\n",
     "observations = df[\"observations\"].tolist()\n",
@@ -972,7 +990,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 33,
+   "execution_count": 34,
    "metadata": {},
    "outputs": [
     {
@@ -981,7 +999,7 @@
        "Text(0.5, 1.0, 'From List')"
       ]
      },
-     "execution_count": 33,
+     "execution_count": 34,
      "metadata": {},
      "output_type": "execute_result"
     },

module03_research_data_in_python/03_05_advanced_numpy.ipynb

@@ -1296,6 +1296,29 @@
    "cell_type": "code",
    "execution_count": 50,
    "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[False,  True,  False, False],\n",
+       "       [False, False,  False, False],\n",
+       "       [False, False,  False,  True]])"
+      ]
+     },
+     "execution_count": 50,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "y_is_one = y == 1\n",
+    "y_is_one"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 51,
+   "metadata": {},
    "outputs": [
     {
      "data": {
@@ -1305,14 +1328,14 @@
        "       [False, False,  True, False]])"
       ]
      },
-     "execution_count": 50,
+     "execution_count": 51,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "iszero = x == y\n",
-    "iszero"
+    "aresame = x == y\n",
+    "aresame"
    ]
   },
   {
@@ -1324,7 +1347,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 51,
+   "execution_count": 52,
    "metadata": {},
    "outputs": [
     {
@@ -1333,13 +1356,13 @@
        "array([ 2,  1, -1, -2, -1,  1])"
       ]
      },
-     "execution_count": 51,
+     "execution_count": 52,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "y[np.logical_not(iszero)]"
+    "y[np.logical_not(aresame)]"
    ]
   },
   {
@@ -1351,16 +1374,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 52,
+   "execution_count": 53,
    "metadata": {},
    "outputs": [],
    "source": [
-    "y[iszero] = 5"
+    "y[aresame] = 5"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 53,
+   "execution_count": 54,
    "metadata": {},
    "outputs": [
     {
@@ -1371,7 +1394,7 @@
        "       [-2, -1,  5,  1]])"
       ]
      },
-     "execution_count": 53,
+     "execution_count": 54,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1391,31 +1414,32 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Numpy memory management can be tricksy:"
+    "NumPy manages memory differently from lists.\n",
+    "Changing an element in a copy of a list does not change the original list."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 54,
+   "execution_count": 55,
    "metadata": {},
    "outputs": [],
    "source": [
-    "x = np.arange(5)\n",
+    "x = list(range(5))\n",
     "y = x[:]"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 55,
+   "execution_count": 56,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "array([0, 1, 0, 3, 4])"
+       "[0, 1, 2, 3, 4]"
       ]
      },
-     "execution_count": 55,
+     "execution_count": 56,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1429,31 +1453,31 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "It does **not** behave like lists!"
+    "But in NumPy, changing the copy **does** change the original array!"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 56,
+   "execution_count": 57,
    "metadata": {},
    "outputs": [],
    "source": [
-    "x = list(range(5))\n",
+    "x = np.arange(5)\n",
     "y = x[:]"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 57,
+   "execution_count": 58,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "[0, 1, 2, 3, 4]"
+       "array([0, 1, 0, 3, 4])"
       ]
      },
-     "execution_count": 57,
+     "execution_count": 58,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1470,13 +1494,6 @@
     "We must use `np.copy` to force separate memory.\n",
     "Otherwise NumPy tries its hardest to make slices be *views* on data."
    ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Now, this has all been very theoretical, but let's go through a practical example, and see how powerful NumPy can be."
-   ]
   }
  ],
  "metadata": {

module03_research_data_in_python/03_06_boids.ipynb

@@ -20,6 +20,8 @@
    "source": [
     "⚠️ Warning: Advanced Topic! ⚠️\n",
     "\n",
+    "Our earlier discussion of NumPy was very theoretical, but let's go through a practical example, and see how powerful NumPy can be.\n",
+    "\n",
     "*Note this is more a showcase of what you can do with numpy than an exhaustive notebook to work through*"
    ]
   },

module03_research_data_in_python/03_07_classroom_exercises.ipynb

@@ -76,6 +76,8 @@
     "\n",
     "The first plot should show sin(x) and cos(x) for the range of x between -1 pi and +1 pi.\n",
     "\n",
+    "**Hint:** The `range(start, stop, step)` function only works with integers. Use the `arange` function from `numpy` instead: `np.arange(start, stop, step)`.\n",
+    "\n",
     "The second plot should show sin(x), cos(x) and the sum of sin(x) and cos(x) over the same -pi to +pi range.\n",
     "Set suitable limits on the axes and pick colours, markers, or line-styles that will make it easy to differentiate between the curves.\n",
     "Add legends to both axes."
@@ -190,7 +192,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.8"
+   "version": "3.8.12"
   }
  },
  "nbformat": 4,