update Roulette docs (#551)

docs/examples/param_space_1d_examples.ipynb

@@ -750,9 +750,9 @@
    "source": [
     "The roulette method allows us to find a prior distribution by *drawing*. The name roulette comes from the analogy that we are placing a limited set of chips where we think the mass of a distribution should be.\n",
     "\n",
-    "For this task we are offered with a grid of m equally sized bins covering the range of x. And we have to allocate a total of n chips between the bins. In other words we use a grid to draw a histogram and the function will try to tell us what distribution, from a given pool of options, is a better fit for our drawing.\n",
+    "For this task we are offered with a grid of `m` equally sized bins covering the range of `x`. And we have to allocate a total of `n` chips between the bins. In other words we use a grid to draw a histogram and the function will try to tell us what distribution, from a given pool of options, is a better fit for our drawing.\n",
     "\n",
-    "The `pz.roulette()` function uses matplotlib interactively, for this to work we need to call the magic `%matplotlib widget` and have ipywidgets installed (a requirement of PreliZ)."
+    "`pz.Roulette()` uses matplotlib interactively, for this to work we need to call the magic `%matplotlib widget` and have ipywidgets installed (a optional requirement of PreliZ). Once you have elicited the distribution you can call `.dist` attribute to get the distribution. In the following example it will be `result.dist`."
    ]
   },
   {
@@ -803,7 +803,7 @@
    ],
    "source": [
     "%matplotlib widget\n",
-    "pz.roulette()"
+    "result = pz.Roulette()"
    ]
   },
   {