liuyu/qlib
1
0
Fork 0
mirror of https://github.com/microsoft/qlib.git synced 2026-07-15 00:36:55 +08:00
Code Activity

Add TabNet config

Browse Source
This commit is contained in:
Jactus
2020-11-25 11:16:01 +08:00
parent 0e2c2fcd7f
commit 991c6195bd
9 changed files with 114 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
qlib/contrib/model/tabnet.py
View File

@@ -9,19 +9,24 @@ from ...model.base import Model
from ...data.dataset import DatasetH
from ...data.dataset.handler import DataHandlerLP
class TabNetModel(Model):
"""TabNetModel Model"""
def __init__(self, n_d, n_a,
n_steps,
gamma,
n_independent,
n_shared,
seed,
momentum,
lambda_sparse,
optimizer_params,
**kwargs):
def __init__(
self,
n_d,
n_a,
n_steps,
gamma,
n_independent,
n_shared,
seed,
momentum,
lambda_sparse,
optimizer_params,
**kwargs
):
self.model = None
self.n_d = n_d
@@ -47,28 +52,28 @@ class TabNetModel(Model):
seed=0,
momentum=0.02,
lambda_sparse=1e-3,
optimizer_params={'lr':2e-3},
optimizer_params={"lr": 2e-3},
**kwargs
):
df_train, df_valid = dataset.prepare(
["train", "valid"], col_set=["feature", "label"], data_key=DataHandlerLP.DK_L
)
x_train, y_train = df_train["feature"].values, df_train["label"].values*100
x_valid, y_valid = df_valid["feature"].values, df_valid["label"].values*100
x_train, y_train = df_train["feature"].values, df_train["label"].values * 100
x_valid, y_valid = df_valid["feature"].values, df_valid["label"].values * 100
self.model = TabNetRegressor(
n_d=self.n_d,
n_a=self.n_a,
n_steps=self.n_steps,
gamma=self.gamma,
n_independent=self.n_independent,
n_shared=self.n_shared,
seed=self.seed,
momentum=self.momentum,
lambda_sparse=self.lambda_sparse,
optimizer_params=self.optimizer_params,
**kwargs
n_d=self.n_d,
n_a=self.n_a,
n_steps=self.n_steps,
gamma=self.gamma,
n_independent=self.n_independent,
n_shared=self.n_shared,
seed=self.seed,
momentum=self.momentum,
lambda_sparse=self.lambda_sparse,
optimizer_params=self.optimizer_params,
**kwargs
)
self.model.fit(x_train, y_train, eval_set=[(x_valid, y_valid)])