Merge pull request #23 from AlbandeCrevoisier/master

Fix typos.

Author: marcoct

tutorials/Introduction to Modeling in Gen.ipynb

@@ -281,7 +281,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "The trace also contains the value of the random choices, stored in map from address to value called a *choice map*. This map is available through the API method [`get_choices`]():"
+    "The trace also contains the value of the random choices, stored in a map from address to value called a *choice map*. This map is available through the API method [`get_choices`]():"
    ]
   },
   {
@@ -796,10 +796,10 @@
     "    end\n",
     "    \n",
     "    # Run the model with new x coordinates, and with parameters \n",
-    "    # fixed to be the inferred values\n",
+    "    # fixed to be the inferred values.\n",
     "    (new_trace, _) = Gen.generate(model, (new_xs,), constraints)\n",
     "    \n",
-    "    # Pull out the y-values and return them\n",
+    "    # Pull out the y-values and return them.\n",
     "    ys = [new_trace[(:y, i)] for i=1:length(new_xs)]\n",
     "    return ys\n",
     "end;"
@@ -911,7 +911,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Now consider the same experiment run with following data set, which has significantly more noise."
+    "Now consider the same experiment run with the following data set, which has significantly more noise."
    ]
   },
   {
@@ -970,7 +970,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Then, we compare the predictions using inference the unmodified and modified model on the `ys` data set:"
+    "Then, we compare the predictions using inference of the unmodified and modified models on the `ys` data set:"
    ]
   },
   {
@@ -998,7 +998,7 @@
    "source": [
     "Notice that there is more uncertainty in the predictions made using the modified model.\n",
     "\n",
-    "We also compare the predictions using inference the unmodified and modified model on the `ys_noisy` data set:"
+    "We also compare the predictions using inference of the unmodified and modified models on the `ys_noisy` data set:"
    ]
   },
   {
@@ -1034,7 +1034,7 @@
     "-------------------------\n",
     "### Exercise\n",
     "\n",
-    "Write a modified version the sine model that makes noise into a random choice. Compare the predicted data with the observed data `infer_and_predict` and `plot_predictions` for the unmodified and modified model, and for the `ys_sine` and `ys_noisy` datasets. Discuss the results. Experiment with the amount of inference computation used. The amount of inference computation will need to be higher for the model with the noise random choice.\n",
+    "Write a modified version of the sine model that makes noise into a random choice. Compare the predicted data with the observed data using `infer_and_predict` and `plot_predictions` for the unmodified and modified models, and for the `ys_sine` and `ys_noisy` data sets. Discuss the results. Experiment with the amount of inference computation used. The amount of inference computation will need to be higher for the model with the noise as a random choice.\n",
     "\n",
     "We have provided you with starter code:"
    ]
@@ -1479,7 +1479,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Gen's built-in modeling language can be used to express models that use an unbounded number of parameters. This section walks you through development of a model of data that does not a-priori specify an upper bound on the complexity of the model, but instead infers the complexity of the model as well as the parameters. This is a simple example of a *Bayesian nonparametric* model."
+    "Gen's built-in modeling language can be used to express models that use an unbounded number of parameters. This section walks you through development of a model of data that does not a priori specify an upper bound on the complexity of the model, but instead infers the complexity of the model as well as the parameters. This is a simple example of a *Bayesian nonparametric* model."
    ]
   },
   {
@@ -1523,7 +1523,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "The data set on the left appears to be best explained as a contant function with some noise. The data set on the right appears to include two changepoints, with a constant function in between the changepoints. We want a model that does not a-priori choose the number of changepoints in the data. To do this, we will recursively partition the interval into regions. We define a Julia data structure that represents a binary tree of intervals; each leaf node represents a region in which the function is constant."
+    "The data set on the left appears to be best explained as a contant function with some noise. The data set on the right appears to include two changepoints, with a constant function in between the changepoints. We want a model that does not a priori choose the number of changepoints in the data. To do this, we will recursively partition the interval into regions. We define a Julia data structure that represents a binary tree of intervals; each leaf node represents a region in which the function is constant."
    ]
   },
   {
@@ -1652,7 +1652,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Now that we have generative function that generates a random piecewise-constant function, we write a model that adds noise to the resulting constant functions to generate a data set of y-coordinates. The noise level will be a random choice."
+    "Now that we have a generative function that generates a random piecewise-constant function, we write a model that adds noise to the resulting constant functions to generate a data set of y-coordinates. The noise level will be a random choice."
    ]
   },
   {
@@ -1677,7 +1677,7 @@
     "    end\n",
     "end\n",
     "\n",
-    "# Out full model\n",
+    "# Our full model\n",
     "@gen function changepoint_model(xs::Vector{Float64})\n",
     "    node = @trace(generate_segments(minimum(xs), maximum(xs)), :tree)\n",
     "    noise = @trace(gamma(1, 1), :noise)\n",