tra-analysis/analysis-master/tra_analysis/SVM.py
Arthur Lu 60beaa4563 tra-analysis v 3.0.0 aggregate PR (#73)
* reflected doc changes to README.md

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* tra_analysis v 2.1.0-alpha.1

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* changed setup.py to use __version__ from source
added Topic and keywords

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* updated Supported Platforms in README.md

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* moved required files back to parent

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* moved security back to parent

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* moved security back to parent
moved contributing back to parent

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* add PR template

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* moved to parent folder

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* moved meta files to .github folder

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* Analysis.py v 3.0.1

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* updated test_analysis for submodules, and added missing numpy import in Sort.py

* fixed item one of Issue #58

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* readded cache searching in postCreateCommand

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* added myself as an author

* feat: created kivy gui boilerplate

* added Kivy to requirements.txt

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* feat: gui with placeholders

* fix: changed config.json path

* migrated docker base image to debian

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* style: spaces to tabs

* migrated to ubuntu

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* fixed issues

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* fix: docker build?

* fix: use ubuntu bionic

* fix: get kivy installed

* @ltcptgeneral can't spell

* optim dockerfile for not installing unused packages

* install basic stuff while building the container

* use prebuilt image for development

* install pylint on base image

* rename and use new kivy

* tests: added tests for Array and CorrelationTest

Both are not working due to errors

* use new thing

* use 20.04 base

* symlink pip3 to pip

* use pip instead of pip3

* equation.Expression.py v 0.0.1-alpha
added corresponding .pyc to .gitignore

* parser.py v 0.0.2-alpha

* added pyparsing to requirements.txt

* parser v 0.0.4-alpha

* Equation v 0.0.1-alpha

* added Equation to tra_analysis imports

* tests: New unit tests for submoduling (#66)

* feat: created kivy gui boilerplate

* migrated docker base image to debian

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* migrated to ubuntu

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* fixed issues

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* fix: docker build?

* fix: use ubuntu bionic

* fix: get kivy installed

* @ltcptgeneral can't spell

* optim dockerfile for not installing unused packages

* install basic stuff while building the container

* use prebuilt image for development

* install pylint on base image

* rename and use new kivy

* tests: added tests for Array and CorrelationTest

Both are not working due to errors

* fix: Array no longer has *args and CorrelationTest functions no longer have self in the arguments

* use new thing

* use 20.04 base

* symlink pip3 to pip

* use pip instead of pip3

* tra_analysis v 2.1.0-alpha.2
SVM v 1.0.1
added unvalidated SVM unit tests

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* fixed version number

Signed-off-by: ltcptgeneral <learthurgo@gmail.com>

* tests: added tests for ClassificationMetric

* partially fixed and commented out svm unit tests

* fixed some SVM unit tests

* added installing pytest to devcontainer.json

* fix: small fixes to KNN

Namely, removing self from parameters and passing correct arguments to KNeighborsClassifier constructor

* fix, test: Added tests for KNN and NaiveBayes.

Also made some small fixes in KNN, NaiveBayes, and RegressionMetric

* test: finished unit tests except for StatisticalTest

Also made various small fixes and style changes

* StatisticalTest v 1.0.1

* fixed RegressionMetric unit test
temporarily disabled CorrelationTest unit tests

* tra_analysis v 2.1.0-alpha.3

* readded __all__

* fix: floating point issues in unit tests for CorrelationTest

Co-authored-by: AGawde05 <agawde05@gmail.com>
Co-authored-by: ltcptgeneral <learthurgo@gmail.com>
Co-authored-by: Dev Singh <dev@devksingh.com>
Co-authored-by: jzpan1 <panzhenyu2014@gmail.com>

* fixed depreciated escape sequences

* ficed tests, indent, import in test_analysis

* changed version to 3.0.0
added backwards compatibility

* ficed pytest install in container

* removed GUI changes

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* incremented version to rc.1 (release candidate 1)

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* fixed NaiveBayes __changelog__

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* fix: __setitem__  == to single =

* Array v 1.0.1

* Revert "Array v 1.0.1"

This reverts commit 59783b79f7.

* Array v 1.0.1

* Array.py v 1.0.2
added more Array unit tests

* cleaned .gitignore
tra_analysis v 3.0.0-rc2

Signed-off-by: Arthur Lu <learthurgo@gmail.com>

* added *.pyc to gitignore
finished subdividing test_analysis

* feat: gui layout + basic func

* Froze and removed superscript (data-analysis)

* remove data-analysis deps install for devcontainer

* tukey pairwise comparison and multicomparison but no critical q-values

* quick patch for devcontainer.json

* better fix for devcontainer.json

* fixed some styling in StatisticalTest
removed print statement in StatisticalTest unit tests

* update analysis tests to be more effecient

* don't use loop for test_nativebayes

* removed useless secondary docker files

* tra-analysis v 3.0.0

Co-authored-by: James Pan <panzhenyu2014@gmail.com>
Co-authored-by: AGawde05 <agawde05@gmail.com>
Co-authored-by: zpan1 <72054510+zpan1@users.noreply.github.com>
Co-authored-by: Dev Singh <dev@devksingh.com>
Co-authored-by: = <=>
Co-authored-by: Dev Singh <dsingh@imsa.edu>
Co-authored-by: zpan1 <zpan@imsa.edu>
2021-04-28 19:33:50 -05:00

88 lines
2.5 KiB
Python

# Titan Robotics Team 2022: SVM submodule
# Written by Arthur Lu
# Notes:
# this should be imported as a python module using 'from tra_analysis import SVM'
# setup:
__version__ = "1.0.2"
__changelog__ = """changelog:
1.0.2:
- fixed __all__
1.0.1:
- removed unessasary self calls
- removed classness
1.0.0:
- ported analysis.SVM() here
"""
__author__ = (
"Arthur Lu <learthurgo@gmail.com>",
)
__all__ = [
"CustomKernel",
"StandardKernel",
"PrebuiltKernel",
"fit",
"eval_classification",
"eval_regression",
]
import sklearn
from sklearn import svm
from . import ClassificationMetric, RegressionMetric
class CustomKernel:
def __new__(cls, C, kernel, degre, gamma, coef0, shrinking, probability, tol, cache_size, class_weight, verbose, max_iter, decision_function_shape, random_state):
return sklearn.svm.SVC(C = C, kernel = kernel, gamma = gamma, coef0 = coef0, shrinking = shrinking, probability = probability, tol = tol, cache_size = cache_size, class_weight = class_weight, verbose = verbose, max_iter = max_iter, decision_function_shape = decision_function_shape, random_state = random_state)
class StandardKernel:
def __new__(cls, kernel, C=1.0, degree=3, gamma='auto_deprecated', coef0=0.0, shrinking=True, probability=False, tol=0.001, cache_size=200, class_weight=None, verbose=False, max_iter=-1, decision_function_shape='ovr', random_state=None):
return sklearn.svm.SVC(C = C, kernel = kernel, gamma = gamma, coef0 = coef0, shrinking = shrinking, probability = probability, tol = tol, cache_size = cache_size, class_weight = class_weight, verbose = verbose, max_iter = max_iter, decision_function_shape = decision_function_shape, random_state = random_state)
class PrebuiltKernel:
class Linear:
def __new__(cls):
return sklearn.svm.SVC(kernel = 'linear')
class Polynomial:
def __new__(cls, power, r_bias):
return sklearn.svm.SVC(kernel = 'polynomial', degree = power, coef0 = r_bias)
class RBF:
def __new__(cls, gamma):
return sklearn.svm.SVC(kernel = 'rbf', gamma = gamma)
class Sigmoid:
def __new__(cls, r_bias):
return sklearn.svm.SVC(kernel = 'sigmoid', coef0 = r_bias)
def fit(kernel, train_data, train_outputs): # expects *2d data, 1d labels or outputs
return kernel.fit(train_data, train_outputs)
def eval_classification(kernel, test_data, test_outputs):
predictions = kernel.predict(test_data)
return ClassificationMetric(predictions, test_outputs)
def eval_regression(kernel, test_data, test_outputs):
predictions = kernel.predict(test_data)
return RegressionMetric(predictions, test_outputs)