- Objectives objective 1. Pull method for expression generation objective 2. Use more graph & algebra et all... objective 3. Build interactive visualization in IPython objective N. Build in stats & logging while dev'n previous 2 Normalize metric values Add random or brownian data to the analysis, should find nothing
- Ipython notebook examples
- D3, or similar
- plot geneology on pareto fronts
- interactive visualizations
- networkx & relations for ...
- algebra
- growing / initing policies
- simplification / expansions
- filtering policies
- +C ???
- OTHER ISSUE:
- dealing with C vs C_&&
- model.orig vs model.expr &&
- init'n vs manip'n
- logging
- statistics
- memotree
- within model
- for expansions
- what's improving and not
- subexpressions- scikit learn
- pandas DFs
- get/set parameters
- pipelining
- gridsearch
- run on the GPU with theano
- distributing to the cloud, pyspark
- diffeqs
- problems with default parameters
- need to toggle on system type ???
- other system types
- invarients
- hidden
- pdes
- abstract expressions / memoization
- when / where coefficients
- domain alphabet
- sub-expression frequencies in population New stuff
- Eqn Relationships, subtree too
- Different metrics for selection & search
- Work around / outside of PGE algorithm (data handling, feature selection, PCA)
- Python package with scikit-learn integration
- More benchmarks, exploring limits
- Memoization