importances = rf.feature_importances_
# included = np.asarray(included)
included = X.columns.values
indices = np.argsort(importances)[::-1]
pf = PlotlyFig(y_title='Importance (%)',
title='Feature by importances',
filename='E:/importances.html',
fontsize=20,
ticksize=15)
pf.bar(x=included[indices][0:10], y=importances[indices][0:10])
#----------------------------------------------------------------
#----------------------------------------------------------------
#----------------------------------------------------------------
pipe = MatPipe.from_preset("express")#the heavy can change to express or light, judge on how exactly the data you want to get
pipe.fit(train_df, target)#this will take a long time
prediction_df = pipe.predict(prediction_df)
prediction_df.to_csv('C:/Users/DELL/Documents/predictionK_VRH.csv')
from sklearn.metrics import mean_absolute_error
from sklearn.dummy import DummyRegressor
# fit the dummy
dr = DummyRegressor()
dr.fit(train_df["composition"], train_df[target])
dummy_test = dr.predict(test_df["composition"])
# Score dummy and MatPipe
true = test_df[target]
matpipe_test = prediction_df[target + " predicted"]
mae_matpipe = mean_absolute_error(true, matpipe_test)
mae_dummy = mean_absolute_error(true, dummy_test)
print("K_VRH Dummy MAE: {} ".format(mae_dummy))
df['Mh'] = mh #df['diel']*df['K_VRH']
df = df.replace([np.inf, -np.inf], np.nan)
df = df.dropna()
df.to_csv('Mh_test.csv')
print(df.describe())
target = 'Mh'
train_df, test_df = train_test_split(df,
test_size=0.1,
shuffle=True,
random_state=1)
prediction_df = test_df.drop(target) #['Mh','K_VRH','diel'],axis=1)
print(prediction_df.columns)
from automatminer import MatPipe
pipe = MatPipe.from_preset("debug", n_jobs=28) #,cache_src='Mh_cache.json')
pipe.fit(train_df, target)
prediction_df = pipe.predict(prediction_df)
from sklearn.metrics import mean_absolute_error
from sklearn.dummy import DummyRegressor
# fit the dummy
dr = DummyRegressor()
dr.fit(train_df["structure"], train_df[target])
dummy_test = dr.predict(test_df["structure"])
# Score dummy and MatPipe
true = test_df[target]
matpipe_test = prediction_df[target + " predicted"]
df = pd.DataFrame(columns=['structure', 'K_VRH'])
df['structure'] = centro_structs
df['K_VRH'] = K_VRH
df = df.dropna()
df.to_csv('centro_elastic.csv')
print(df.describe())
train_df, test_df = train_test_split(df,
test_size=0.1,
shuffle=True,
random_state=1)
target = "K_VRH"
prediction_df = test_df.drop(columns=[target])
pipe = MatPipe.from_preset("express")
pipe.fit(train_df, target)
prediction_df = pipe.predict(prediction_df)
# fit the dummy
dr = DummyRegressor()
dr.fit(train_df["structure"], train_df[target])
dummy_test = dr.predict(test_df["structure"])
# Score dummy and MatPipe
true = test_df[target]
matpipe_test = prediction_df[target + " predicted"]
mae_matpipe = mean_absolute_error(true, matpipe_test)
mse_matpipe = mean_squared_error(true, matpipe_test)
df = pd.DataFrame(columns=['structure', 'dielectric'])
df['structure'] = centro_structs
df['dielectric'] = diel
df = df.dropna()
df.to_csv('centro_diel.csv')
print(df.describe())
train_df, test_df = train_test_split(df,
test_size=0.1,
shuffle=True,
random_state=1)
target = "dielectric"
prediction_df = test_df.drop(columns=[target])
pipe = MatPipe.from_preset("express", n_jobs=28, cache_src="cache_diel.json")
pipe.fit(train_df, target)
prediction_df = pipe.predict(prediction_df)
# fit the dummy
dr = DummyRegressor()
dr.fit(train_df["structure"], train_df[target])
dummy_test = dr.predict(test_df["structure"])
# Score dummy and MatPipe
true = test_df[target]
matpipe_test = prediction_df[target + " predicted"]
mae_matpipe = mean_absolute_error(true, matpipe_test)
mse_matpipe = mean_squared_error(true, matpipe_test)