This no longer works.
The s3 data copies are no longer there.

Test Notebook
In [1]:  import bokeh
In [2]:  bokeh.__version__
Out[2]: u'1.3.4'
In [3]: import blaze
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-4-cf28b9f4f41e> in <module>()
----> 1 import blaze

/Users/dap/opt/anaconda3/envs/py27/lib/python2.7/site-packages/blaze/__init__.py in <module>()
      7 
      8 from pandas import DataFrame
----> 9 from odo import odo, convert, append, drop, resource
     10 from odo.backends.csv import CSV
     11 from odo.backends.json import JSON, JSONLines

/Users/dap/opt/anaconda3/envs/py27/lib/python2.7/site-packages/odo/__init__.py in <module>()
     27     from .backends.sas import sas7bdat
     28 with ignoring(ImportError):
---> 29     from .backends.pandas import pd
     30 with ignoring(ImportError):
     31     from .backends.bcolz import bcolz

/Users/dap/opt/anaconda3/envs/py27/lib/python2.7/site-packages/odo/backends/pandas.py in <module>()
    100 
    101 
--> 102 @convert.register((pd.Timestamp, pd.Timedelta), (pd.tslib.NaTType, type(None)))
    103 def convert_null_or_nat_to_nat(n, **kwargs):
    104     return pd.NaT

AttributeError: 'module' object has no attribute 'tslib'

This is with a python=2.7 env I created. Does not work with python=3.8 or 3.7.

I think it would've been helpful to have a backchannel to share code snippets, ask questions, etc. without disrupting the tutorial. I've seen the Mozilla Etherpad used quite successfully in tutorials like this before.

Hi Christine!

Thank you for the time and effort you put into preparing the tutorial!

I had a suggestion for modifying the tutorial material post-tutorial session, hopefully this doesn't take up too much time, and hopefully you might consider it.

Would you consider putting URLs to the pages where we could grab hints/learn by examples? A few examples:

1. Notebook (NB) 1.2: prior to "Axis Format", a link to a page that shows an example of formatting axis dates.
2. NB 1.2: a link to the figure() source code, so that we can inspect the figure attributes.

There are others too that would have helped us during the coding challenges.

Hope you have a great rest of the conference!

Thank you Christine very much for sharing this. Any chance this will be updated for the newer Bokeh/Pandas releases. I have tried today to follow the video and the first two notebooks and found out they will not run under Bokeh 11.1. TimeSeries for example does not take an "index" argument; now there is an "x" argument. Even Pandas 0.17.1 does not accept an "index" argument in its read_csv function; I had to use the method "set_index" on the loaded data frame to get it to work. And the list goes on. So what say you? It would be a shame to leave this work to go to waste after such a very short time, wouldn't you agree?

Hi, less an issue here and more commentary.

I very much enjoyed your presentation. I found the flow from simple plots to more extensive, integrated dashboards was excellent. I was tripped up by some small things, and thought the following might improve the whole for any future tutorials:

• Avoid file reading with NetCDF, or provide the code to do the reading directly. The tutorial is about Bokeh, and the focus should not stray from it.
• Build/aggregate reusable code blocks more directly. Your cells with comments in the notebook are very helpful for breaking the problem up into parts, but it was frustrating to say "OK, now copy-paste your earlier code into this new area (but go fetch your mess of code from lots of cells earlier.)" It was disorienting at best. I don't know a good compromise between keeping the nice in-line display and problem break-up of the IPython Notebook, while also keeping a more file-based view that is easy to change and work into cohesive code blocks. So, I'm whining without a suggestion of a solution (sorry), but wanted to mention this as something that slowed down my learning.
• Show the complete example (the animation from elsewhere) at the very beginning, and use that as motivation for the entire day's activities. I had a much greater appreciation for the different parts of the tutorial once I realized how you were tying them together.

Thanks again for your hard work on this tutorial.