def generate_dataset(x1, x2, y):
#load data
qm9_tasks, datasets, transformers = load_qm9()
train_dataset, valid_dataset, test_dataset = datasets
print("x1 = ", qm9_tasks[x1 - 1])
print("x2 = ", qm9_tasks[x2 - 1])
print("y = ", qm9_tasks[y - 1])
#extrct the 'y'values
Y = test_dataset.y
YT = Y.T
X1 = YT[x1 - 1]
X2 = YT[x2 - 1]
Y_a = YT[y - 1]
x1 = X1.tolist()
x2 = X2.tolist()
y_l = Y_a.tolist()
l = Y_a.shape
n = np.random.uniform(0, l,
1).astype(np.int) #set the number of noise added
ni = np.random.uniform(0, l, n) #n random values
an = len(ni)
#add noise to n numbers of y
for i in range(an):
mu = 0
sigma = (x1[i] + x2[i]) / 2
noise = np.random.normal(mu, np.abs(sigma), n)
g = noise.tolist()
y_l[i] += g[i]
#save to_csv:
dataframe = pd.DataFrame({'x1': X1, 'x2': X2, 'y': y_l})
dataframe.to_csv("y_gen_data.csv", index=False, sep=',')
gen_data = pd.read_csv('y_gen_data.csv')
"""
Script that trains Tensorflow multitask models on QM9 dataset.
"""
from __future__ import print_function
from __future__ import division
from __future__ import unicode_literals
import os
import deepchem as dc
import numpy as np
from deepchem.molnet import load_qm9
np.random.seed(123)
qm9_tasks, datasets, transformers = load_qm9()
train_dataset, valid_dataset, test_dataset = datasets
fit_transformers = [dc.trans.CoulombFitTransformer(train_dataset)]
regression_metric = [
dc.metrics.Metric(dc.metrics.mean_absolute_error, mode="regression"),
dc.metrics.Metric(dc.metrics.pearson_r2_score, mode="regression")
]
model = dc.models.MultiTaskFitTransformRegressor(
n_tasks=len(qm9_tasks),
n_features=[29, 29],
learning_rate=0.001,
momentum=.8,
batch_size=32,
weight_init_stddevs=[1 / np.sqrt(400), 1 / np.sqrt(100), 1 / np.sqrt(100)],
bias_init_consts=[0., 0., 0.],
layer_sizes=[400, 100, 100],
dropouts=[0.01, 0.01, 0.01],
fit_transformers=fit_transformers,