Written by by Justin B. Kinney, Cold Spring Harbor Laboratory
Last updated on 23 March 2015
Reference: Kinney JB (2014) Unification of Field Theory and Maximum Entropy Methods for Learning Probability Densities. arXiv:1411.5371 [physics.data-an] http://arxiv.org/abs/1411.5371
Code: https://github.com/jbkinney/14_maxent
=== Instructions ===
All computations were performed using the Canopy Python Environment by Enthought, available at https://www.enthought.com/products/canopy/
Dependencies: scipy
The file deft_nobc.py and deft_utils.py contain all the density estmation routines used in this paper. Just copy this file to any directory in which you want to peform density estimation, and do execute "from deft_nobc.py import deft_nobc_1d" in your Python script
To see a simple demsontration of the function "deft_nobc_1d" in 1 dimension, run demo.py
To recreate Fig. 2, run fig_2.py.
To recreate Fig. 3, run fig_3_calculate.py, then fig_3_draw.py.
To recreate Fig. 4, run fig_4.py.
=== FILES ===
demo.py: Contains a quick example of DEFT without boundary conditions in 1D.
deft_nobc.py: Contains primary functions for field theory density estimation in 1D without boundary conditions.
deft_utils.py Contains helper functions for field theory density estimation in 1D without boundary conditions.
fig_2.py: Performs computations for and plots Fig. 2. Saves as fig_2.pdf
fig_3_calculate.py: Performs computations for Fig. 3. Saves results in fig_3_20.pickle (for 20 simulations, not 100).
fig_3_draw.py: Draws Fig. 3, saves as fig_3.pdf
fig_3.pickle: Contains results of simulations performed for fig_3.pdf
fig_4.py: Draws Fig. 4, saves as fig_4.pdf