update general docs with new tutorial

Author: phobson

docs/index.rst

@@ -61,6 +61,7 @@ Tutorial
    :maxdepth: 2
 
    tutorial/getting_started.rst
+   tutorial/closer_look_at_viz.rst
 
 Testing
 -------

docs/tutorial/Makefile

@@ -1,3 +1,4 @@
 notebooks:
 
-	tools/nb_to_doc.py getting_started
+	tools/nb_to_doc.py getting_started  
+	tools/nb_to_doc.py closer_look_at_viz

docs/tutorial/getting_started.ipynb

@@ -8,15 +8,25 @@
     "\n",
     "## Installation\n",
     "\n",
+    "`mpl-probscale` is developed on Python 3.5. It is also tested on Python 3.4 and even 2.7 (for the time being).\n",
+    "\n",
     "### From conda\n",
-    "I've uploaded the package to my anaconda channel.\n",
+    "Official releases of `mpl-probscale` can be found on conda-forge:\n",
+    "\n",
+    "`conda install --channel=conda-forge mpl-probscale`\n",
+    "\n",
+    "Fairly recent builds of the development verions are available on my channel:\n",
     "\n",
     "`conda install --channel=phobson mpl-probscale`\n",
     "\n",
-    "### From pip\n",
+    "\n",
+    "### From PyPI\n",
+    "\n",
+    "Official source releases are also available on PyPI\n",
     "`pip install probscale`\n",
     "\n",
     "### From source\n",
+    "\n",
     "`mpl-probscale` is a pure python package. It should be fairly trivial to install from source on any platform. To do that, download or clone from [github](https://github.com/phobson/mpl-probscale), unzip the archive if necessary then do:\n",
     "\n",
     "```\n",

probscale/viz.py

@@ -96,7 +96,7 @@ def probplot(data, ax=None, axtype='prob', dist=None, probax='x',
 
     # compute the plotting positions and sort the data
     probs, datavals = plot_pos(data, **pp_kws)
-    qntls = _minimal_norm.ppf(probs)
+    qntls = dist.ppf(probs)
 
     # determine how the probability values should be expressed
     if axtype == 'qq':