# stop = None
# This random_state is under Linux system. For others system ,the random_state maybe different。
# try with different random_state.
stop = lambda ind: ind.fitness.values[0] >= 0.99
sl = SymbolLearning(loop='MultiMutateLoop',
pset=pset0,
gen=10,
pop=3000,
hall=1,
batch_size=40,
re_hall=5,
n_jobs=12,
mate_prob=0.8,
max_value=h_bgp,
initial_min=1,
initial_max=h_bgp,
mutate_prob=0.8,
tq=True,
dim_type="coef",
stop_condition=stop,
re_Tree=0,
random_state=0,
verbose=True,
add_coef=True,
inter_add=True,
cal_dim=False,
inner_add=True,
personal_map=False)
# sl.fit()
print(sl.expr)
pset0.add_constants(c, c_dim=c_dim, c_prob=0.05)
pset0.add_operations(power_categories=(2, 3, 0.5, 1 / 3, 4, 1 / 4),
# categories=("Mul",),
categories=("Add", "Mul", "Sub", "Div", "exp", "ln"),
self_categories=None)
total_height = 3
h_bgp = 2
# This random_state is under Linux system. For others system ,the random_state maybe different,please
# try with different random_state.
for i in range(1, 10):
stop = lambda ind: ind.fitness.values[0] >= 0.95
sl = SymbolLearning(loop="MultiMutateLoop", pset=pset0, gen=20, pop=1000, hall=1, batch_size=40, re_hall=3,
n_jobs=12, mate_prob=0.9, max_value=h_bgp, initial_min=2, initial_max=h_bgp,
mutate_prob=0.8, tq=False, dim_type="coef", stop_condition=stop,
re_Tree=0, store=False, random_state=4, verbose=True,
# stats=None,
stats={"fitness_dim_max": ["max"], "dim_is_target": ["sum"], "h_bgp": ["mean"]},
add_coef=True, inter_add=True, out_add=True, cal_dim=True, vector_add=True,
personal_map=False)
tt.t
sl.fit()
tt.t
tt.p
score = sl.score(x, y, "r2")
print(sl.expr)
y_pre = sl.predict(x)
break
# just for shown
# y_pre = si_transformer.scale_y * y_pre
# ssc = Dim.inverse_convert(y_dim, target_units=eV)[0]
c = [1, 2, 3]
# data = pd.read_csv(r"..data/a.csv")
# y = data[:,0]
# x = data[:,1:]
sl = SymbolLearning(
loop="MultiMutateLoop",
pop=500,
re_hall=3,
gen=5,
random_state=1,
# classification=True,
classification=False,
# scoring=[metrics.accuracy_score, ], score_pen=[1, ],
store=False,
# store=True,
n_jobs=4,
batch_size=20,
batch_para=False,
# add_coef=False,
# out_add =True,
# vector_add=True,
)
tt.t
sl.fit(
x,
y,
c=c,
# x_group=[[0, 1], [2, 3]]
# @Time : 2020/12/16 18:53 # @Email : [email protected] # @Software: PyCharm # @License: BSD 3-Clause from sklearn import metrics from sklearn.utils import shuffle if __name__ == "__main__": from sklearn.datasets import fetch_california_housing from bgp.skflow import SymbolLearning data = fetch_california_housing() x = data["data"][:100] y = data["target"][:100] sl = SymbolLearning(loop="MultiMutateLoop", pop=50, gen=2, random_state=1) sl.fit(x, y) score = sl.score(x, y, "r2") print(sl.expr) from sklearn.datasets import load_iris from bgp.skflow import SymbolLearning data = load_iris() x = data["data"][:98, :] x[40:60] = shuffle(x[40:60], random_state=2) y = data["target"][:98] c = None sl = SymbolLearning(loop="MultiMutateLoop", pop=50,
est_gp = SymbolLearning(
loop='MultiMutateLoop',
pop=1000,
gen=20,
mutate_prob=0.5,
mate_prob=0.8,
hall=1,
re_hall=1,
re_Tree=None,
initial_min=None,
initial_max=3,
max_value=4,
scoring=(metrics.mean_absolute_error, ),
score_pen=(-1, ),
filter_warning=True,
cv=1,
add_coef=True,
inter_add=True,
inner_add=True,
vector_add=False,
out_add=False,
flat_add=False,
cal_dim=False,
dim_type=None,
fuzzy=False,
n_jobs=8,
batch_size=40,
random_state=1,
stats=None,
verbose=True,
migrate_prob=0,
tq=True,
store=True,
personal_map=False,
stop_condition=None,
details=False,
classification=False,
score_object="y",
)
if __name__ == "__main__":
from sklearn.datasets import fetch_california_housing
from bgp.skflow import SymbolLearning
data = fetch_california_housing()
x = data["data"][:100]
y = data["target"][:100]
sl = SymbolLearning(loop="MultiMutateLoop", pop=500, gen=2, random_state=1)
sl.fit(x, y, x_group=[[1, 2], [3, 4], [6, 7]])
score = sl.score(x, y, "r2")
print(sl.expr)
# classification=True, scoring=[metrics.accuracy_score, ], score_pen=[1, ],
# cal_dim=True, n_jobs = 10, store=True,
#
# )
# # 方式选择3，系数加在公式内层,认定系数可以自动补全量纲
pset0.y_dim = None
sl = SymbolLearning(
loop="MultiMutateLoop",
pop=1000,
gen=20,
random_state=1,
pset=pset0,
classification=True,
scoring=[
metrics.accuracy_score,
],
score_pen=[
1,
],
cal_dim=True,
inner_add=True,
out_add=False,
n_jobs=10,
)
#
# 方式选择4，系数加在公式内层,不考虑量纲计算,
# sl = SymbolLearning(loop="MultiMutateLoop", pop=200, gen=10, random_state=1,pset=pset0,
# classification=True, scoring=[metrics.accuracy_score, ], score_pen=[1, ],
# cal_dim=False,
# inner_add=False, out_add=False,
# n_jobs = 2,)
from sklearn.metrics import r2_score, mean_absolute_error
dataTi = pd.read_csv("Ti.csv", index_col=0)
datanp = dataTi.values
x = datanp[:, 1:]
y = datanp[:, 0]
ind = [0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
index = np.where(np.array(ind) == 1)[0]
x = x[:, index]
sl = SymbolLearning(loop='MultiMutateLoop', pop=1000, gen=10, random_state=1, n_jobs=1,
add_coef=True, initial_max=2, initial_min=1,
max_value=2, store=False, )
tt.t
sl.fit(x, y)
tt.t
tt.p
print(sl.expr)
print(sl.fitness)
# y_pre = sl.predict(x)
x0 = x[:, 0]
x1 = x[:, 1]
y_pre2 = x1 / x0 ** 3
s1 = r2_score(y, y_pre2)
lr = LinearRegression()
###########第一个###########
"""数据"""
com_data = pd.read_csv(r'FCC.csv')
x = com_data.iloc[:, :-1].values
y = com_data.iloc[:, -1].values
x, y = shuffle(x, y, random_state=0)
st = Store("FCC_result_error_no_intercept")
st.start()
sl = SymbolLearning(loop=r'MultiMutateLoop',
cal_dim=False,
dim_type=pa_dim,
pop=5000,
gen=50,
add_coef=True,
re_hall=2,
inter_add=False,
random_state=2,
n_jobs=16,
initial_max=2,
max_value=4,
store=True,
stats={"fitness_dim_max": ("max", )})
sl.fit(
x,
y,
x_dim=[pa_dim, dless, dless, dless, dless, dless, dless],
y_dim=pa_dim,
power_categories=(2, 3, 0.5, 0.33),
categories=("Add", "Mul", "Sub", "Div"),
)
x, y = shuffle(x, y, random_state=0)
st = Store("reg3_result_error")
st.start()
sl = SymbolLearning(loop=r'MultiMutateLoop',
cal_dim=True,
dim_type=None,
pop=2000,
gen=30,
add_coef=True,
re_hall=2,
scoring=[
error,
],
score_pen=[
-1,
],
inter_add=False,
random_state=3,
n_jobs=30,
initial_max=3,
max_value=6,
store=True,
stats={
"fitness_dim_min": ("min", ),
"dim_is_target": ("sum", )
})
sl.fit(
x,
y,
x_dim=[dless, pa_dim, dless, dless, dless, dless, dless, dless],