mgraffg / evodag Goto Github PK

Naivebayes and naivebayesMN

Repeated code

Set property in model classes

Hi!

I am trying to use EvoDAGE however I fail. I would like to use it as a library.
Trying something like the following:
`from EvoDAG.model import EvoDAGE
from sklearn.datasets import load_iris

data = load_iris()
X = data.data
y = data.target

m = EvoDAGE(n_estimators=30, n_jobs=1).fit(X, y)`

Howver, I receive that error:

ModuleNotFoundError: No module named 'EvoDAG.linalg_solve'

Maybe I have not installed all the required modules(?).
I would really appreciate it if you could help me.

Thank you in advance.

Let us use the fact that
\prod x ^ \alpha = \log \sum \alpha x

Currently, EvoDAG can handle multiple outputs, so it is desirable to tackle multi-label problems

It is using mask instead of mask_ts

in EvoDAGE, it might be useful for AutoML

set log loss as fitness

Set a parameter in params and train to select the fitness function.

Bug when NaiveBayesMN or NaiveBayes is used as input

use n_jobs in predict

This method should be called inside when the dependent variable is set

All ensembles must be built using EvoDAGE.

This parameter should be store in the parameters file.

The idea is to evolve DAG where all the functions are bitwise operations.

(venv) amolina@keken:~/repo/EvoMSA$ pip install EvoDAG
Collecting EvoDAG
Using cached https://files.pythonhosted.org/packages/6c/b3/65112b3ae7683bbe56bf15cc0c425c98d7c49ba6476824f67d0fd94b9536/EvoDAG-0.16.0.tar.gz
ERROR: Complete output from command python setup.py egg_info:
ERROR: /home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Compiler/Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: /tmp/pip-install-nqz9ogww/EvoDAG/EvoDAG/cython_utils.pyx
tree = Parsing.p_module(s, pxd, full_module_name)
warning: EvoDAG/cython_utils.pyx:72:22: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:74:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:75:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:75:33: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:75:50: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:124:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:124:61: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:124:77: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:151:22: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:153:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:154:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:154:33: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:188:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:188:58: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:590:22: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:590:26: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:591:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/cython_utils.pyx:591:26: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
/home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Compiler/Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: /tmp/pip-install-nqz9ogww/EvoDAG/EvoDAG/linalg_solve.pyx
tree = Parsing.p_module(s, pxd, full_module_name)
warning: EvoDAG/linalg_solve.pyx:66:16: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/linalg_solve.pyx:66:30: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
warning: EvoDAG/linalg_solve.pyx:111:21: Non-trivial type declarators in shared declaration (e.g. mix of pointers and values). Each pointer declaration should be on its own line.
/home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Compiler/Main.py:367: FutureWarning: Cython directive 'language_level' not set, using 2 for now (Py2). This will change in a later release! File: /tmp/pip-install-nqz9ogww/EvoDAG/EvoDAG/function_selection.pyx
tree = Parsing.p_module(s, pxd, full_module_name)

Error compiling Cython file:
------------------------------------------------------------
...
        return False

    cpdef int tournament(self) except -1:
        cdef int best, comp
        cdef double best_fit, comp_fit
        cdef unsigned int *nargs = self.nargs.data.as_uints
                                                 ^
------------------------------------------------------------

EvoDAG/function_selection.pyx:85:50: Cannot convert Python object to 'unsigned int *'
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/tmp/pip-install-nqz9ogww/EvoDAG/setup.py", line 63, in <module>
    'boundscheck': False}),
  File "/home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Build/Dependencies.py", line 1097, in cythonize
    cythonize_one(*args)
  File "/home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Build/Dependencies.py", line 1220, in cythonize_one
    raise CompileError(None, pyx_file)
Cython.Compiler.Errors.CompileError: EvoDAG/function_selection.pyx
Compiling EvoDAG/linalg_solve.pyx because it depends on /home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/SparseArray/sparse_array.pxd.
Compiling EvoDAG/function_selection.pyx because it depends on /home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/Cython/Includes/cpython/weakref.pxd.
Compiling EvoDAG/cython_utils.pyx because it depends on /home/amolina/repo/EvoMSA/venv/lib/python3.6/site-packages/SparseArray/sparse_array.pxd.
[1/3] Cythonizing EvoDAG/cython_utils.pyx
[2/3] Cythonizing EvoDAG/linalg_solve.pyx
[3/3] Cythonizing EvoDAG/function_selection.pyx
----------------------------------------

ERROR: Command "python setup.py egg_info" failed with error code 1 in /tmp/pip-install-nqz9ogww/EvoDAG/
(venv) amolina@keken:~/repo/EvoMSA$ pip3 freeze
b4msa==2.1
Cython==0.29.7
microtc==2.1.1
numpy==1.16.3
SparseArray==0.8.15

Handle NaN on inputs and outputs

set the function set in node.py

It is necessary to set a time limit, per run, for example one hr.

In order to evolve classifiers based on prototypes, EvoDAG requieres to implement strongly type GP strategies.