These are some IPython Notebooks to get started with the Connectomic Challenge at Kaggle.
The code shows how to download and import the datasets and does a few pre-processing steps such as determining signaling distance matrices.
The most important part is probably the spike detector, and the subsequent visualization of possible causal influences on neuron spike probabilities.
I conclude with a short draft for formulas of a Bayesian Model.
Also included in this set of IPython Notebooks is example code of how to create and embed HTML 5 animations in IPython Notebooks using Matplotlib and FFMPEG.
Please click on the following links to directly view these IPython Notebooks via nbviewer.ipython.org, as they should be.
- Notebook 1: The Connectomics Data Set
- [Notebook 2: Spikes and their Detection] (http://nbviewer.ipython.org/github/kadeng/connectomics-ipython/blob/master/2%20-%20Connectomics%20Spike%20Detection.ipynb)
- [Notebook 3: Visualizing and modeling causal influences] (http://nbviewer.ipython.org/github/kadeng/connectomics-ipython/blob/master/3%20-%20Connectomics%20-%20a%20Peek%20at%20Correlations%20and%20Influences.ipynb)
- Supplement: How to create animations using Matplotlib and FFMPEG