Quantum Machinery with gdb1k



In [ ]:

    
%load_ext autoreload
%autoreload 2
%pdb off
__author__ = "Joseph Gomes and Bharath Ramsundar"
__copyright__ = "Copyright 2016, Stanford University"
__license__ = "LGPL"

import os
import unittest

import numpy as np
import deepchem as dc
import numpy.random
from deepchem.utils.evaluate import Evaluator
from sklearn.ensemble import RandomForestRegressor
from sklearn.kernel_ridge import KernelRidge

Setting up model variables



In [ ]:

    
featurizer = dc.feat.CoulombMatrixEig(23, remove_hydrogens=False)
tasks = ["atomization_energy"]
dataset_file = "../../datasets/gdb1k.sdf"
smiles_field = "smiles"
mol_field = "mol"

Load featurized data



In [ ]:

    
loader = dc.data.SDFLoader(
      tasks=["atomization_energy"], smiles_field="smiles",
      featurizer=featurizer,
      mol_field="mol")
dataset = loader.featurize(dataset_file)

Perform Train, Validation, and Testing Split



In [ ]:

    
random_splitter = dc.splits.RandomSplitter()
train_dataset, valid_dataset, test_dataset = random_splitter.train_valid_test_split(dataset)

Transforming datasets



In [ ]:

    
transformers = [
    dc.trans.NormalizationTransformer(transform_X=True, dataset=train_dataset),
    dc.trans.NormalizationTransformer(transform_y=True, dataset=train_dataset)]

for dataset in [train_dataset, valid_dataset, test_dataset]:
  for transformer in transformers:
      dataset = transformer.transform(dataset)

Fit Random Forest with hyperparameter search



In [ ]:

    
def rf_model_builder(model_params, model_dir):
  sklearn_model = RandomForestRegressor(**model_params)
  return dc.models.SklearnModel(sklearn_model, model_dir)
params_dict = {
    "n_estimators": [10, 100],
    "max_features": ["auto", "sqrt", "log2", None],
}

metric = dc.metrics.Metric(dc.metrics.mean_absolute_error)
optimizer = dc.hyper.HyperparamOpt(rf_model_builder)
best_rf, best_rf_hyperparams, all_rf_results = optimizer.hyperparam_search(
    params_dict, train_dataset, valid_dataset, transformers,
    metric=metric)



In [ ]:

    
def krr_model_builder(model_params, model_dir):
  sklearn_model = KernelRidge(**model_params)
  return dc.models.SklearnModel(sklearn_model, model_dir)

params_dict = {
    "kernel": ["laplacian"],
    "alpha": [0.0001],
    "gamma": [0.0001]
}

metric = dc.metrics.Metric(dc.metrics.mean_absolute_error)
optimizer = dc.hyper.HyperparamOpt(krr_model_builder)
best_krr, best_krr_hyperparams, all_krr_results = optimizer.hyperparam_search(
    params_dict, train_dataset, valid_dataset, transformers,
    metric=metric)