Update documentation on seed

Author: Henri Vuollekoski

tutorial.ipynb

@@ -36,7 +36,7 @@
     "%precision 2\n",
     "\n",
     "# Set an arbitrary seed and a global random state to keep the randomly generated quantities the same between runs\n",
-    "seed = 20170530\n",
+    "seed = 20170530  # this will be separately given to ELFI\n",
     "np.random.seed(seed)"
    ]
   },
@@ -644,7 +644,7 @@
     "\n",
     "A common optional keyword argument, accepted by all inference methods, `batch_size` defines how many simulations are performed in each passing through the graph. \n",
     "\n",
-    "Another optional keyword is the seed. This ensures that the outcome will be always the same for the same data and model. If you leave it out, a random seed will be taken. Giving `seed='global'` will use the global numpy seed."
+    "Another optional keyword is the seed. This ensures that the outcome will be always the same for the same data and model. If you leave it out, a random seed will be taken."
    ]
   },
   {