From 71b5a6af9fcebf4163bc4cf2553c4b6a1d68c268 Mon Sep 17 00:00:00 2001 From: danijoo Date: Sun, 19 Dec 2021 11:23:39 +0100 Subject: [PATCH] finish Excercise 1 --- Exercise1/exercise1.ipynb | 788 ++++++++++++++++++++++++++++++++------ 1 file changed, 675 insertions(+), 113 deletions(-) diff --git a/Exercise1/exercise1.ipynb b/Exercise1/exercise1.ipynb index 38a9fae..7cec9ae 100755 --- a/Exercise1/exercise1.ipynb +++ b/Exercise1/exercise1.ipynb @@ -19,7 +19,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 1, "metadata": {}, "outputs": [], "source": [ @@ -108,7 +108,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 2, "metadata": {}, "outputs": [], "source": [ @@ -126,7 +126,7 @@ " Return the 5x5 identity matrix.\n", " \"\"\" \n", " # ======== YOUR CODE HERE ======\n", - " A = [] # modify this line\n", + " A = np.eye(5)\n", " \n", " # ==============================\n", " return A" @@ -149,9 +149,24 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 3, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "text/plain": [ + "array([[1., 0., 0., 0., 0.],\n", + " [0., 1., 0., 0., 0.],\n", + " [0., 0., 1., 0., 0.],\n", + " [0., 0., 0., 1., 0.],\n", + " [0., 0., 0., 0., 1.]])" + ] + }, + "execution_count": 3, + "metadata": {}, + "output_type": "execute_result" + } + ], "source": [ "warmUpExercise()" ] @@ -173,9 +188,31 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 4, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Normal Equations | 10 / 10 | Nice work!\n", + " Warm up exercise | 0 / 40 | Your answer is incorrect.\n", + " Computing Cost (for one variable) | 0 / 50 | Your answer is incorrect.\n", + " Gradient Descent (for one variable) | 0 / 0 | Your answer is incorrect.\n", + " Feature Normalization | 0 / 0 | Your answer is incorrect.\n", + " Computing Cost (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " Gradient Descent (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 10 / 100 | \n", + "\n" + ] + } + ], "source": [ "# appends the implemented function in part 1 to the grader object\n", "grader[1] = warmUpExercise\n", @@ -199,7 +236,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 5, "metadata": {}, "outputs": [], "source": [ @@ -236,7 +273,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 6, "metadata": {}, "outputs": [], "source": [ @@ -268,11 +305,12 @@ " can also set the marker edge color using the `mec` property.\n", " \"\"\"\n", " fig = pyplot.figure() # open a new figure\n", - " \n", - " # ====================== YOUR CODE HERE ======================= \n", - " \n", - "\n", - " # =============================================================\n" + " pyplot.scatter(x, y)\n", + " pyplot.gca().set(\n", + " xlabel = \"Profit / $10k\",\n", + " ylabel = \"Population / 10k\"\n", + " )\n", + "\n" ] }, { @@ -288,9 +326,22 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 7, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "plotData(X, y)" ] @@ -306,9 +357,288 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 8, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\u001b[1;31mSignature:\u001b[0m \u001b[0mpyplot\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mscalex\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mscaley\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mTrue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n", + "\u001b[1;31mDocstring:\u001b[0m\n", + "Plot y versus x as lines and/or markers.\n", + "\n", + "Call signatures::\n", + "\n", + " plot([x], y, [fmt], *, data=None, **kwargs)\n", + " plot([x], y, [fmt], [x2], y2, [fmt2], ..., **kwargs)\n", + "\n", + "The coordinates of the points or line nodes are given by *x*, *y*.\n", + "\n", + "The optional parameter *fmt* is a convenient way for defining basic\n", + "formatting like color, marker and linestyle. It's a shortcut string\n", + "notation described in the *Notes* section below.\n", + "\n", + ">>> plot(x, y) # plot x and y using default line style and color\n", + ">>> plot(x, y, 'bo') # plot x and y using blue circle markers\n", + ">>> plot(y) # plot y using x as index array 0..N-1\n", + ">>> plot(y, 'r+') # ditto, but with red plusses\n", + "\n", + "You can use `.Line2D` properties as keyword arguments for more\n", + "control on the appearance. Line properties and *fmt* can be mixed.\n", + "The following two calls yield identical results:\n", + "\n", + ">>> plot(x, y, 'go--', linewidth=2, markersize=12)\n", + ">>> plot(x, y, color='green', marker='o', linestyle='dashed',\n", + "... linewidth=2, markersize=12)\n", + "\n", + "When conflicting with *fmt*, keyword arguments take precedence.\n", + "\n", + "\n", + "**Plotting labelled data**\n", + "\n", + "There's a convenient way for plotting objects with labelled data (i.e.\n", + "data that can be accessed by index ``obj['y']``). Instead of giving\n", + "the data in *x* and *y*, you can provide the object in the *data*\n", + "parameter and just give the labels for *x* and *y*::\n", + "\n", + ">>> plot('xlabel', 'ylabel', data=obj)\n", + "\n", + "All indexable objects are supported. This could e.g. be a `dict`, a\n", + "`pandas.DataFame` or a structured numpy array.\n", + "\n", + "\n", + "**Plotting multiple sets of data**\n", + "\n", + "There are various ways to plot multiple sets of data.\n", + "\n", + "- The most straight forward way is just to call `plot` multiple times.\n", + " Example:\n", + "\n", + " >>> plot(x1, y1, 'bo')\n", + " >>> plot(x2, y2, 'go')\n", + "\n", + "- Alternatively, if your data is already a 2d array, you can pass it\n", + " directly to *x*, *y*. A separate data set will be drawn for every\n", + " column.\n", + "\n", + " Example: an array ``a`` where the first column represents the *x*\n", + " values and the other columns are the *y* columns::\n", + "\n", + " >>> plot(a[0], a[1:])\n", + "\n", + "- The third way is to specify multiple sets of *[x]*, *y*, *[fmt]*\n", + " groups::\n", + "\n", + " >>> plot(x1, y1, 'g^', x2, y2, 'g-')\n", + "\n", + " In this case, any additional keyword argument applies to all\n", + " datasets. Also this syntax cannot be combined with the *data*\n", + " parameter.\n", + "\n", + "By default, each line is assigned a different style specified by a\n", + "'style cycle'. The *fmt* and line property parameters are only\n", + "necessary if you want explicit deviations from these defaults.\n", + "Alternatively, you can also change the style cycle using the\n", + "'axes.prop_cycle' rcParam.\n", + "\n", + "\n", + "Parameters\n", + "----------\n", + "x, y : array-like or scalar\n", + " The horizontal / vertical coordinates of the data points.\n", + " *x* values are optional and default to `range(len(y))`.\n", + "\n", + " Commonly, these parameters are 1D arrays.\n", + "\n", + " They can also be scalars, or two-dimensional (in that case, the\n", + " columns represent separate data sets).\n", + "\n", + " These arguments cannot be passed as keywords.\n", + "\n", + "fmt : str, optional\n", + " A format string, e.g. 'ro' for red circles. See the *Notes*\n", + " section for a full description of the format strings.\n", + "\n", + " Format strings are just an abbreviation for quickly setting\n", + " basic line properties. All of these and more can also be\n", + " controlled by keyword arguments.\n", + "\n", + " This argument cannot be passed as keyword.\n", + "\n", + "data : indexable object, optional\n", + " An object with labelled data. If given, provide the label names to\n", + " plot in *x* and *y*.\n", + "\n", + " .. note::\n", + " Technically there's a slight ambiguity in calls where the\n", + " second label is a valid *fmt*. `plot('n', 'o', data=obj)`\n", + " could be `plt(x, y)` or `plt(y, fmt)`. In such cases,\n", + " the former interpretation is chosen, but a warning is issued.\n", + " You may suppress the warning by adding an empty format string\n", + " `plot('n', 'o', '', data=obj)`.\n", + "\n", + "Other Parameters\n", + "----------------\n", + "scalex, scaley : bool, optional, default: True\n", + " These parameters determined if the view limits are adapted to\n", + " the data limits. The values are passed on to `autoscale_view`.\n", + "\n", + "**kwargs : `.Line2D` properties, optional\n", + " *kwargs* are used to specify properties like a line label (for\n", + " auto legends), linewidth, antialiasing, marker face color.\n", + " Example::\n", + "\n", + " >>> plot([1,2,3], [1,2,3], 'go-', label='line 1', linewidth=2)\n", + " >>> plot([1,2,3], [1,4,9], 'rs', label='line 2')\n", + "\n", + " If you make multiple lines with one plot command, the kwargs\n", + " apply to all those lines.\n", + "\n", + " Here is a list of available `.Line2D` properties:\n", + "\n", + " agg_filter: a filter function, which takes a (m, n, 3) float array and a dpi value, and returns a (m, n, 3) array\n", + " alpha: float\n", + " animated: bool\n", + " antialiased or aa: bool\n", + " clip_box: `.Bbox`\n", + " clip_on: bool\n", + " clip_path: [(`~matplotlib.path.Path`, `.Transform`) | `.Patch` | None]\n", + " color or c: color\n", + " contains: callable\n", + " dash_capstyle: {'butt', 'round', 'projecting'}\n", + " dash_joinstyle: {'miter', 'round', 'bevel'}\n", + " dashes: sequence of floats (on/off ink in points) or (None, None)\n", + " drawstyle or ds: {'default', 'steps', 'steps-pre', 'steps-mid', 'steps-post'}, default: 'default'\n", + " figure: `.Figure`\n", + " fillstyle: {'full', 'left', 'right', 'bottom', 'top', 'none'}\n", + " gid: str\n", + " in_layout: bool\n", + " label: object\n", + " linestyle or ls: {'-', '--', '-.', ':', '', (offset, on-off-seq), ...}\n", + " linewidth or lw: float\n", + " marker: marker style\n", + " markeredgecolor or mec: color\n", + " markeredgewidth or mew: float\n", + " markerfacecolor or mfc: color\n", + " markerfacecoloralt or mfcalt: color\n", + " markersize or ms: float\n", + " markevery: None or int or (int, int) or slice or List[int] or float or (float, float)\n", + " path_effects: `.AbstractPathEffect`\n", + " picker: float or callable[[Artist, Event], Tuple[bool, dict]]\n", + " pickradius: float\n", + " rasterized: bool or None\n", + " sketch_params: (scale: float, length: float, randomness: float)\n", + " snap: bool or None\n", + " solid_capstyle: {'butt', 'round', 'projecting'}\n", + " solid_joinstyle: {'miter', 'round', 'bevel'}\n", + " transform: `matplotlib.transforms.Transform`\n", + " url: str\n", + " visible: bool\n", + " xdata: 1D array\n", + " ydata: 1D array\n", + " zorder: float\n", + "\n", + "Returns\n", + "-------\n", + "lines\n", + " A list of `.Line2D` objects representing the plotted data.\n", + "\n", + "See Also\n", + "--------\n", + "scatter : XY scatter plot with markers of varying size and/or color (\n", + " sometimes also called bubble chart).\n", + "\n", + "Notes\n", + "-----\n", + "**Format Strings**\n", + "\n", + "A format string consists of a part for color, marker and line::\n", + "\n", + " fmt = '[marker][line][color]'\n", + "\n", + "Each of them is optional. If not provided, the value from the style\n", + "cycle is used. Exception: If ``line`` is given, but no ``marker``,\n", + "the data will be a line without markers.\n", + "\n", + "Other combinations such as ``[color][marker][line]`` are also\n", + "supported, but note that their parsing may be ambiguous.\n", + "\n", + "**Markers**\n", + "\n", + "============= ===============================\n", + "character description\n", + "============= ===============================\n", + "``'.'`` point marker\n", + "``','`` pixel marker\n", + "``'o'`` circle marker\n", + "``'v'`` triangle_down marker\n", + "``'^'`` triangle_up marker\n", + "``'<'`` triangle_left marker\n", + "``'>'`` triangle_right marker\n", + "``'1'`` tri_down marker\n", + "``'2'`` tri_up marker\n", + "``'3'`` tri_left marker\n", + "``'4'`` tri_right marker\n", + "``'s'`` square marker\n", + "``'p'`` pentagon marker\n", + "``'*'`` star marker\n", + "``'h'`` hexagon1 marker\n", + "``'H'`` hexagon2 marker\n", + "``'+'`` plus marker\n", + "``'x'`` x marker\n", + "``'D'`` diamond marker\n", + "``'d'`` thin_diamond marker\n", + "``'|'`` vline marker\n", + "``'_'`` hline marker\n", + "============= ===============================\n", + "\n", + "**Line Styles**\n", + "\n", + "============= ===============================\n", + "character description\n", + "============= ===============================\n", + "``'-'`` solid line style\n", + "``'--'`` dashed line style\n", + "``'-.'`` dash-dot line style\n", + "``':'`` dotted line style\n", + "============= ===============================\n", + "\n", + "Example format strings::\n", + "\n", + " 'b' # blue markers with default shape\n", + " 'or' # red circles\n", + " '-g' # green solid line\n", + " '--' # dashed line with default color\n", + " '^k:' # black triangle_up markers connected by a dotted line\n", + "\n", + "**Colors**\n", + "\n", + "The supported color abbreviations are the single letter codes\n", + "\n", + "============= ===============================\n", + "character color\n", + "============= ===============================\n", + "``'b'`` blue\n", + "``'g'`` green\n", + "``'r'`` red\n", + "``'c'`` cyan\n", + "``'m'`` magenta\n", + "``'y'`` yellow\n", + "``'k'`` black\n", + "``'w'`` white\n", + "============= ===============================\n", + "\n", + "and the ``'CN'`` colors that index into the default property cycle.\n", + "\n", + "If the color is the only part of the format string, you can\n", + "additionally use any `matplotlib.colors` spec, e.g. full names\n", + "(``'green'``) or hex strings (``'#008000'``).\n", + "\u001b[1;31mFile:\u001b[0m c:\\users\\danijoo\\anaconda3\\lib\\site-packages\\matplotlib\\pyplot.py\n", + "\u001b[1;31mType:\u001b[0m function\n" + ] + } + ], "source": [ "?pyplot.plot" ] @@ -352,7 +682,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 9, "metadata": {}, "outputs": [], "source": [ @@ -377,7 +707,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 10, "metadata": {}, "outputs": [], "source": [ @@ -416,12 +746,7 @@ " m = y.size # number of training examples\n", " \n", " # You need to return the following variables correctly\n", - " J = 0\n", - " \n", - " # ====================== YOUR CODE HERE =====================\n", - "\n", - " \n", - " # ===========================================================\n", + " J = 1/(2*m) * np.power((np.dot(theta, X.T) - y), 2).sum() \n", " return J" ] }, @@ -434,9 +759,23 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 11, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "With theta = [0, 0] \n", + "Cost computed = 32.07\n", + "Expected cost value (approximately) 32.07\n", + "\n", + "With theta = [-1, 2]\n", + "Cost computed = 54.24\n", + "Expected cost value (approximately) 54.24\n" + ] + } + ], "source": [ "J = computeCost(X, y, theta=np.array([0.0, 0.0]))\n", "print('With theta = [0, 0] \\nCost computed = %.2f' % J)\n", @@ -457,9 +796,31 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 12, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Normal Equations | 10 / 10 | Nice work!\n", + " Warm up exercise | 40 / 40 | Nice work!\n", + " Computing Cost (for one variable) | 0 / 50 | Your answer is incorrect.\n", + " Gradient Descent (for one variable) | 0 / 0 | Your answer is incorrect.\n", + " Feature Normalization | 0 / 0 | Your answer is incorrect.\n", + " Computing Cost (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " Gradient Descent (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 50 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[2] = computeCost\n", "grader.grade()" @@ -494,7 +855,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 73, "metadata": {}, "outputs": [], "source": [ @@ -543,16 +904,22 @@ " # are passed by reference to functions\n", " theta = theta.copy()\n", " \n", - " J_history = [] # Use a python list to save cost in every iteration\n", - " \n", + " J_history = np.zeros(num_iters) # save cost in every iteration here\n", + " print(X.shape)\n", " for i in range(num_iters):\n", - " # ==================== YOUR CODE HERE =================================\n", - " \n", - "\n", - " # =====================================================================\n", + " theta_tmp = theta.copy()\n", + " for j in range(len(theta)):\n", + " # Vectorized steepest gradient step\n", + " # evaluate the hypothesis for all data sets\n", + " hypothesis = np.tensordot(X, theta, axes=(1, 0))\n", + " # calculate the sum ( H(x) - y * x(i) )\n", + " sum = ((hypothesis - y) * X[:, j]).sum()\n", + " # update theta\n", + " theta_tmp[j] = theta[j] - alpha/m * sum\n", " \n", " # save the cost J in every iteration\n", - " J_history.append(computeCost(X, y, theta))\n", + " J_history[i] = computeCost(X, y, theta)\n", + " theta = theta_tmp\n", " \n", " return theta, J_history" ] @@ -566,9 +933,18 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 14, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Theta found by gradient descent: -3.6303, 1.1664\n", + "Expected theta values (approximately): [-3.6303, 1.1664]\n" + ] + } + ], "source": [ "# initialize fitting parameters\n", "theta = np.zeros(2)\n", @@ -593,9 +969,22 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 15, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "# plot the linear fit\n", "plotData(X[:, 1], y)\n", @@ -620,9 +1009,20 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 16, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "For population = 35,000, we predict a profit of 4519.77\n", + "\n", + "For population = 70,000, we predict a profit of 45342.45\n", + "\n" + ] + } + ], "source": [ "# Predict values for population sizes of 35,000 and 70,000\n", "predict1 = np.dot([1, 3.5], theta)\n", @@ -641,9 +1041,31 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 17, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Normal Equations | 10 / 10 | Nice work!\n", + " Warm up exercise | 40 / 40 | Nice work!\n", + " Computing Cost (for one variable) | 50 / 50 | Nice work!\n", + " Gradient Descent (for one variable) | 0 / 0 | Your answer is incorrect.\n", + " Feature Normalization | 0 / 0 | Your answer is incorrect.\n", + " Computing Cost (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " Gradient Descent (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 100 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[3] = gradientDescent\n", "grader.grade()" @@ -666,9 +1088,22 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 18, "metadata": {}, - "outputs": [], + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "# grid over which we will calculate J\n", "theta0_vals = np.linspace(-10, 10, 100)\n", @@ -730,9 +1165,28 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 83, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + " X[:,0] X[:, 1] y\n", + "--------------------------\n", + " 2104 3 399900\n", + " 1600 3 329900\n", + " 2400 3 369000\n", + " 1416 2 232000\n", + " 3000 4 539900\n", + " 1985 4 299900\n", + " 1534 3 314900\n", + " 1427 3 198999\n", + " 1380 3 212000\n", + " 1494 3 242500\n" + ] + } + ], "source": [ "# Load data\n", "data = np.loadtxt(os.path.join('Data', 'ex1data2.txt'), delimiter=',')\n", @@ -773,7 +1227,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 84, "metadata": {}, "outputs": [], "source": [ @@ -810,14 +1264,10 @@ " You might find the 'np.mean' and 'np.std' functions useful.\n", " \"\"\"\n", " # You need to set these values correctly\n", + " mu = X.mean(axis=0)\n", + " sigma = X.std(axis=0)\n", " X_norm = X.copy()\n", - " mu = np.zeros(X.shape[1])\n", - " sigma = np.zeros(X.shape[1])\n", - "\n", - " # =========================== YOUR CODE HERE =====================\n", - "\n", - " \n", - " # ================================================================\n", + " X_norm = (X_norm-mu)/sigma\n", " return X_norm, mu, sigma" ] }, @@ -830,15 +1280,26 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 85, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Computed mean: [2000.68085106 3.17021277]\n", + "Computed standard deviation: [7.86202619e+02 7.52842809e-01]\n", + "Min/Max values: [-1.46104938 -2.88269044] [3.15099326 2.4305037 ]\n" + ] + } + ], "source": [ "# call featureNormalize on the loaded data\n", "X_norm, mu, sigma = featureNormalize(X)\n", "\n", "print('Computed mean:', mu)\n", - "print('Computed standard deviation:', sigma)" + "print('Computed standard deviation:', sigma)\n", + "print('Min/Max values:', X_norm.min(axis=0), X_norm.max(axis=0))" ] }, { @@ -850,9 +1311,32 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 36, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + "(20, 3)\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Normal Equations | 10 / 10 | Nice work!\n", + " Warm up exercise | 40 / 40 | Nice work!\n", + " Computing Cost (for one variable) | 50 / 50 | Nice work!\n", + " Gradient Descent (for one variable) | 0 / 0 | Nice work!\n", + " Feature Normalization | 0 / 0 | Your answer is incorrect.\n", + " Computing Cost (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " Gradient Descent (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 100 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[4] = featureNormalize\n", "grader.grade()" @@ -867,7 +1351,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 86, "metadata": {}, "outputs": [], "source": [ @@ -912,7 +1396,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 87, "metadata": {}, "outputs": [], "source": [ @@ -945,13 +1429,8 @@ " m = y.shape[0] # number of training examples\n", " \n", " # You need to return the following variable correctly\n", - " J = 0\n", - " \n", - " # ======================= YOUR CODE HERE ===========================\n", - "\n", - " \n", - " # ==================================================================\n", - " return J\n" + " y_est = np.tensordot(X, theta, axes=(1, 0))\n", + " return 1/(2*m) * (np.power((y_est - y), 2)).sum()\n" ] }, { @@ -963,9 +1442,33 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 75, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + "(20, 3)\n", + "(20, 4)\n", + " Part Name | Score | Feedback\n", + " --------- | ----- | --------\n", + " Normal Equations | 10 / 10 | Nice work!\n", + " Warm up exercise | 40 / 40 | Nice work!\n", + " Computing Cost (for one variable) | 50 / 50 | Nice work!\n", + " Gradient Descent (for one variable) | 0 / 0 | Nice work!\n", + " Feature Normalization | 0 / 0 | Nice work!\n", + " Computing Cost (for multiple variables) | 0 / 0 | Nice work!\n", + " Gradient Descent (for multiple variables) | 0 / 0 | Your answer is incorrect.\n", + " --------------------------------\n", + " | 100 / 100 | \n", + "\n" + ] + } + ], "source": [ "grader[5] = computeCostMulti\n", "grader.grade()" @@ -980,7 +1483,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 90, "metadata": {}, "outputs": [], "source": [ @@ -1021,24 +1524,8 @@ " While debugging, it can be useful to print out the values of \n", " the cost function (computeCost) and gradient here.\n", " \"\"\"\n", - " # Initialize some useful values\n", - " m = y.shape[0] # number of training examples\n", - " \n", - " # make a copy of theta, which will be updated by gradient descent\n", - " theta = theta.copy()\n", - " \n", - " J_history = []\n", - " \n", - " for i in range(num_iters):\n", - " # ======================= YOUR CODE HERE ==========================\n", - "\n", - " \n", - " # =================================================================\n", - " \n", - " # save the cost J in every iteration\n", - " J_history.append(computeCostMulti(X, y, theta))\n", - " \n", - " return theta, J_history" + " # Single variable solution also works with multiple dimensions. So I reuse it here ... \n", + " return gradientDescent(X, y, theta, alpha, num_iters)" ] }, { @@ -1050,11 +1537,43 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 100, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(20, 4)\n" + ] + }, + { + "data": { + "text/plain": [ + "[]" + ] + }, + "execution_count": 100, + "metadata": {}, + "output_type": "execute_result" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "grader[6] = gradientDescentMulti\n", + "X, y, theta, alpha, steps = grader.X2, grader.Y2, np.array([-0.1, -0.2, -0.3, -0.4]), 0.01, 10\n", + "theta, J = gradientDescentMulti(X, y, theta, 0.01, 100)\n", "grader.grade()" ] }, @@ -1094,9 +1613,31 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 106, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "(20, 4)\n", + "theta computed from gradient descent: [0.43662103 0.97758074 0.86398339 0.63363385]\n", + "Predicted price of a 1650 sq-ft, 3 br house (using gradient descent): $0\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], "source": [ "\"\"\"\n", "Instructions\n", @@ -1119,11 +1660,11 @@ "At prediction, make sure you do the same feature normalization.\n", "\"\"\"\n", "# Choose some alpha value - change this\n", - "alpha = 0.1\n", + "alpha = 0.01\n", "num_iters = 400\n", "\n", "# init theta and run gradient descent\n", - "theta = np.zeros(3)\n", + "theta = np.zeros(4)\n", "theta, J_history = gradientDescentMulti(X, y, theta, alpha, num_iters)\n", "\n", "# Plot the convergence graph\n", @@ -1171,7 +1712,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 7, "metadata": {}, "outputs": [], "source": [ @@ -1194,7 +1735,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 2, "metadata": {}, "outputs": [], "source": [ @@ -1224,13 +1765,7 @@ " ----\n", " Look up the function `np.linalg.pinv` for computing matrix inverse.\n", " \"\"\"\n", - " theta = np.zeros(X.shape[1])\n", - " \n", - " # ===================== YOUR CODE HERE ============================\n", - "\n", - " \n", - " # =================================================================\n", - " return theta" + " return np.linalg.inv(X.T@X)@X.T@y \n" ] }, { @@ -1242,9 +1777,20 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 5, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\n", + "Submitting Solutions | Programming Exercise linear-regression\n", + "\n", + "Invalid email or token. You used an invalid email or your token may have expired. Please make sure you have entered all fields correctly. Try generating a new token if the issue still persists.\n" + ] + } + ], "source": [ "grader[7] = normalEqn\n", "grader.grade()" @@ -1262,9 +1808,18 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 8, "metadata": {}, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Theta computed from the normal equations: [89597.9095428 139.21067402 -8738.01911233]\n", + "Predicted price of a 1650 sq-ft, 3 br house (using normal equations): $0\n" + ] + } + ], "source": [ "# Calculate the parameters from the normal equation\n", "theta = normalEqn(X, y);\n", @@ -1281,6 +1836,13 @@ "\n", "print('Predicted price of a 1650 sq-ft, 3 br house (using normal equations): ${:.0f}'.format(price))" ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] } ], "metadata": { @@ -1299,7 +1861,7 @@ "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", - "version": "3.6.6" + "version": "3.7.6" } }, "nbformat": 4,