Python Jupyter Notebooks

I prefer to use Jupyter Notebooks to gather and clean datasets (mostly using pandas and numpy libraries). For me, using python to explore data and create visualizations is less intuitive than using R, so I prefer R for exploratory analysis and visualization. These are new skills for me so I'm curious to hear how folks with more experience approach the analysis process. Thanks!

Kyle Wilbert Cool.  I'd be interested in hearing more.  Right now I'm working on two things.

1. Methods for getting bulk data out of Tessitura and other resources.
   1. How you are ingesting your data.  That is how do you get your data out of the Tessitura Database and other sources into your Notebooks?  Are you using? .CSV or .XLSX files?  Or do you have another approach?
   2. Over the past month or so, I've been working on automating data export from Tessitura. I have some preliminary code working in R and Python for accessing the Tessitura REST API.
2. Once you have to data in R or Python.  What kinds of things are you doing with the data?
   1. As a museum, for example, are you doing and attendance forecasting?
   2. I've been doing some work in that area recently.

--Tom

1. I've primarily been using csv exports. I've used Python to download Twitter info via API - I haven't tried it with Tessitura yet. I'd be interested in seeing what you've got so far.

2. I haven't done any forecasting yet. It's on my list - I'm still getting up to speed with the math and stats behind it. I'm working on a project now in a Python Notebook merging our School Programs data with some NYC data on school (location and demographics). Basically just trying to see who we're engaging and where they're from. Predictive analytics is further in the horizon for me at the moment. In the meantime, we've been demoing Dexibit which has some predictive models included in its library.

1. I've primarily been using csv exports. I've used Python to download Twitter info via API - I haven't tried it with Tessitura yet. I'd be interested in seeing what you've got so far.

2. I haven't done any forecasting yet. It's on my list - I'm still getting up to speed with the math and stats behind it. I'm working on a project now in a Python Notebook merging our School Programs data with some NYC data on school (location and demographics). Basically just trying to see who we're engaging and where they're from. Predictive analytics is further in the horizon for me at the moment. In the meantime, we've been demoing Dexibit which has some predictive models included in its library.

Kyle Wilbert what is the best way to share a Notebook with you?  I've got notebooks using both R and Python to get to data from the V15 REST API. There is even some talk among the Tessitura Developer community about building a Python Tessitura Library.  Try `pip install tessitura` and you can get to a very early version that may still be a bit buggy.

When it comes to forecasting you might want to take a look at `Facebook Prophet` this works on either R or Python.  

I'm really interested in hearing more about what you are doing with the NYC school demographic data.   Are you using data from NYC open Data?  If so what do you think is the best data source there?

Dexibit looks really interesting.  Love to learn more about what you are doing with them.

Hi Kyle - How are you finding Dexibit? AWMM has been on a journey with them for a while now, but as we're still bringing different system data into Dexibit we've yet to start looking at the forecast/predictive modelling. 

Would love to catch up offline if you don't mind to see how your org is finding the system? If you're happy to chat my email is jbramley@aucklandmuseum.com

Jo