#labels.append('heatRelease')
labels.append('T')
labels.append('PVs')
# # tabulate psi, mu, alpha
# labels.append('psi')
# labels.append('mu')
# labels.append('alpha')
# DO NOT CHANGE THIS ORDER!!
input_features = ['f', 'zeta', 'pv']
# read in the data
X, y, df, in_scaler, out_scaler = read_h5_data('./data/tables_of_fgm.h5',
input_features=input_features,
labels=labels,
i_scaler='no',
o_scaler='cbrt_std')
#('./data/tables_of_fgm.h5',key='of_tables',
# in_labels=input_features, labels = labels,scaler=scaler)
# split into train and test data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.01)
# %%
print('set up ANN')
# ANN parameters
dim_input = X_train.shape[1]
dim_label = y_train.shape[1]
input_features = ["f", "zeta", "pv"]
with open("GRI_species_order", "r") as f:
labels = [a.rstrip() for a in f.readlines()]
# append other fields: heatrelease, T, PVs
# labels.append('heatRelease')
labels.append("T")
labels.append("PVs")
#%%
x, y, df, in_scaler, out_scaler = read_h5_data(
"./data/tables_of_fgm.h5",
input_features=input_features,
labels=labels,
i_scaler="no",
o_scaler="cbrt_std",
)
# %%
zetaLevel = list(set(df.zeta))
df_sample = df[df.zeta == zetaLevel[0]].sample(n=5_000)
sp = "T"
px.scatter_3d(data_frame=df_sample,
x="f",
y="pv",
z=sp,
color=sp,
width=800,
# labels.append('heatRelease')
labels.append("T")
labels.append("PVs")
labels.remove("N2")
# # tabulate psi, mu, alpha
# labels.append('psi')
# labels.append('mu')
# labels.append('alpha')
# read in the data
X, y, df, in_scaler, out_scaler = read_h5_data(
# "./data/tables_of_fgm.h5",
# "./data/df_filtered_3.parquet",
"./data/df_interpolation.parquet",
input_features=input_features,
labels=labels,
i_scaler="no",
o_scaler="cbrt_std",
)
# split into train and test data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.01)
# %%
print("set up ANN")
# ANN parameters
dim_input = X_train.shape[1]
dim_label = y_train.shape[1]
# print(species)
labels = f.read().splitlines()
labels.append('T')
labels.append('PVs')
print('The labels are:')
print(labels)
# DO NOT CHANGE THIS ORDER!!
input_features=['f','zeta','pv']
# read in the data
X, y, df, in_scaler, out_scaler = read_h5_data(path_to_data,
input_features=input_features,
labels=labels,
i_scaler='std2',
o_scaler=o_scaler)
# split into train and test data
test_size=data_points/len(X)
print('Test size is %f of entire data set\n' % test_size)
X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=test_size)
# load the model
model = load_model(path_to_model)
# #############################
# inference part
t_start = time.time()
predict_val = model.predict(X_test)