Draft intro page by fonnesbeck · Pull Request #2393 · pymc-devs/pymc

fonnesbeck · 2017-07-06T21:49:30Z

No description provided.

aseyboldt

I wrote down some random stuff, feel free to ignore if you don't agree (english isn't even my native language after all ;)

aseyboldt · 2017-07-06T21:58:27Z

docs/source/intro.rst

+:Web site: https://github.com/pymc-devs/pymc3
+
+
+Purpose


I don't think we need a section title. It's kind of obvious that we introduce pymc here :-)

aseyboldt · 2017-07-06T22:18:31Z

docs/source/intro.rst

+Purpose
+=======
+
+PyMC3 is a probabilistic programming module for Python that allows users to fit Bayesian models using a variety of numerical methods, most notably, Markov chain Monte Carlo (MCMC) and variational inference (VI). Its flexibility and extensibility make it applicable to a large suite of problems. Along with core model specification and fitting functionality, PyMC3 includes methods for summarizing output, plotting, goodness-of-fit and model diagnostics.


The rules for commas in English are a bit of a mystery to me, but do we need the comma after "most notably"?
"suite of problems" sounds a bit strange to me (but there are a lot of google hits, so maybe that is a thing after all). How about something like "It it being used in biology, astronomy, physics, the social sciences, business and finance and can be extended to fit new needs" (we might want to check that list however). I guess that would cover "flexibility".
Maybe "... fitting functionality, PyMC3 can visualize its results and provides model diagnostics."?

aseyboldt · 2017-07-06T22:19:38Z

docs/source/intro.rst

+Features
+========
+
+PyMC3 strives to make Bayesian modeling as simple and painless as possible, thereby allowing users to focus on their scientific problem, rather than on the methods used to solve it. Here is a partial list of its features:


"thereby allowing" -> "to allow"?
I don't think we need the second sentence. Clear from context.

aseyboldt · 2017-07-06T22:21:29Z

docs/source/intro.rst

+
+PyMC3 strives to make Bayesian modeling as simple and painless as possible, thereby allowing users to focus on their scientific problem, rather than on the methods used to solve it. Here is a partial list of its features:
+
+* Fits Bayesian statistical models using modern techniques, including MCMC and VI.


Third person sounds a bit strange. Maybe imperative? Or just "Modern techniques for fitting Bayesian stats models..."
Replace MCMC by hamiltonian MCMC and no abbreviations.

aseyboldt · 2017-07-06T22:24:05Z

docs/source/intro.rst

+
+* Uses Theano as the computational backend, allowing for fast expression evaluation, automatic gradient calculation, and GPU computing.
+
+* Includes a module for Gaussian process modeling.


"Build-in support for Gaussian process modeling"

aseyboldt · 2017-07-06T22:25:42Z

docs/source/intro.rst

+
+* Creates summaries including tables and plots.
+
+* Several convergence diagnostics are available.


"Model checking and convergence detection"

aseyboldt · 2017-07-06T22:28:02Z

docs/source/intro.rst

+  with the full power of Python.
+
+
+What's new in version 3


Mabe just mention that it is a rewrite, but not much detail. That could be on a separate page pymc2 vs pymc3. I could be very wrong about this, but so far I didn't have the impression that most pymc3 users also knew pymc2. For those users the comparison is kind of pointless. Also, this kind of ends up repeating the feature list from above.
The important part of this is I think mentioned in the history section.

I've tried to restrict this list to differences. We do get a fair number of users moving over, and probably will get more now that v.3 is released.

aseyboldt · 2017-07-06T22:29:21Z

docs/source/intro.rst

+
+First, import the PyMC3 functions and classes you will need for building your model. You can import the entire module via `import pymc3 as pm`, or just bring in what you need::
+
+    from pymc3 import Model, Normal, invlogit, Binomial, sample, traceplot


Pretty much everywhere we are using import pymc3 as pm. So I think we should do the same thing here.

Both are valid. I've noted the option for pm above.

aseyboldt · 2017-07-06T22:30:16Z

docs/source/intro.rst

+        theta = invlogit(alpha + beta*dose)
+
+        # Model likelihood
+        deaths = Binomial('deaths', n=n, p=theta, observed=np.array([0, 1, 3, 5]))


Move the definition of the observations before the model def?

aseyboldt · 2017-07-06T22:30:36Z

docs/source/intro.rst

+
+Save this file, then from a python shell (or another file in the same directory), call::
+
+	with bioassay_model:


The indentation is strange

I don't se a problem locally. I will double check.

fonnesbeck · 2017-07-07T00:20:11Z

Thanks! I've integrated several of these suggestions.

aseyboldt · 2017-08-04T19:02:51Z

@fonnesbeck Is this finished?

fonnesbeck · 2017-08-04T21:47:00Z

Yeah, I think so. It can be improved with subsequent PRs if desired.

fonnesbeck added 2 commits July 6, 2017 16:48

Draft intro page

b53a578

Fixed header in intro

30ee412

aseyboldt reviewed Jul 6, 2017

View reviewed changes

fonnesbeck added 10 commits July 6, 2017 19:22

Updated text with suggested edits

02244bc

Converted tabs to spaces throughout

89993d6

Reorganized sections in intro

1723b42

Minor edits to intro

5dea602

Fixed some bad rst

e4fd502

Added all data to example

a44fa97

Added image with example in intro

69160db

Edited text in intro

83fcb78

Fixed typo

202dd8c

Added intro to index

cf3b577

junpenglao merged commit 6e04322 into master Aug 5, 2017

junpenglao deleted the intro_doc branch August 5, 2017 05:47

		:Web site: https://github.com/pymc-devs/pymc3


		Purpose


		PyMC3 strives to make Bayesian modeling as simple and painless as possible, thereby allowing users to focus on their scientific problem, rather than on the methods used to solve it. Here is a partial list of its features:

		* Fits Bayesian statistical models using modern techniques, including MCMC and VI.


		* Uses Theano as the computational backend, allowing for fast expression evaluation, automatic gradient calculation, and GPU computing.

		* Includes a module for Gaussian process modeling.


		* Creates summaries including tables and plots.

		* Several convergence diagnostics are available.


		First, import the PyMC3 functions and classes you will need for building your model. You can import the entire module via `import pymc3 as pm`, or just bring in what you need::

		from pymc3 import Model, Normal, invlogit, Binomial, sample, traceplot


		Save this file, then from a python shell (or another file in the same directory), call::

		with bioassay_model:

Conversation

fonnesbeck commented Jul 6, 2017

Uh oh!

aseyboldt left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aseyboldt Jul 6, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

fonnesbeck commented Jul 7, 2017

Uh oh!

aseyboldt commented Aug 4, 2017

Uh oh!

fonnesbeck commented Aug 4, 2017

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

aseyboldt Jul 6, 2017 •

edited

Loading