def test_score():
opt=Optimizer(loss='mse',optimizer='bfgs',regularizer='ridge',lam=0.01,
problem_type='classification')
opt.fit(X2,y2,iters=1000, verbose=True)
a = opt.predict_proba(X2)[0]
assert a>=0
def test_opty_classification():
opt=Optimizer(loss='mse',optimizer='bfgs',regularizer='ridge',lam=0.01,
problem_type='classification')
opt.fit(X2,y2,iters=1000, verbose=True)
assert opt.coefs[0] <= 3
def test_score2():
opt=Optimizer(loss='mse',optimizer='sgd',regularizer='ridge',lam=0.01,
problem_type='regression')
opt.fit(X2,y2,iters=1000, verbose=True)
a = opt.score(X2,y2)
assert a>=0
def test_opty_verbose_steepestgd():
opt=Optimizer(loss='mse',optimizer='steepestgd',lam=0.01)
opt.fit(X2,y2,iters=1000, verbose=True)
assert opt.coefs[0] <= 3
def test_opty_verbose_bfgs():
opt=Optimizer(loss='mse',optimizer='bfgs',regularizer='ridge',lam=0.01)
opt.fit(X2,y2,iters=1000, verbose=True)
assert opt.coefs[0] <= 3
def test_opty_verbose_gd():
opt=Optimizer(loss=user_mse,optimizer='gd',lam=0.01)
opt.fit(X2,y2,iters=1000, verbose=True)
assert opt.coefs[0] <= 3
def test_get_loss2():
opt1 = Optimizer(loss='mse',optimizer='gd',regularizer='lasso',lam=0.01)
with pytest.raises(Exception):
opt1._get_loss_fn('yo')
def test_opty4():
opt=Optimizer(loss=user_mse,optimizer='gd',lam=0.01)
opt.fit(X2,y2,iters=1000)
assert opt.coefs[0] <= 3
def test_opty2():
opt=Optimizer(loss=user_mse,optimizer='sgd',regularizer='ridge',lam=0.01)
opt.fit(X2,y2,iters=1000)
assert opt.coefs[0] <= 3
def test_opty1():
opt=Optimizer(loss='mse',optimizer='gd',regularizer='lasso',lam=0.01)
opt.fit(X2,y2,iters=1000)
assert opt.coefs[0] <= 3
def test_cost():
opt1 = Optimizer(loss='mse',optimizer='gd',regularizer='lasso',lam=0.01)
assert type(opt1._get_cost_fn('mse', optimizer='gd')).__name__ == "method"
def test_get_loss3():
with pytest.raises(Exception):
opt2=Optimizer(loss='mse',optimizer='homie',regularizer='lasso',lam=0.01)
opt2._get_loss_fn('homie')