Updates for final draft

import shutil

import deepchem as dc

import pandas as pd

import matplotlib

# Force matplotlib to not use any Xwindows backend.

matplotlib.use('Agg')

import matplotlib.mlab as mlab

import matplotlib.pyplot as plt

from FACTORS_datasets import load_factors

###Load data###

np.random.seed(123)

shard_size = 2000

print("About to load FACTORS data.")

FACTORS_tasks, datasets, transformers = load_factors(shard_size=shard_size)

all_results = []

for task in range(n_tasks):

  y_task = y_train[:, task]

  task_results = []

  for other_task in range(n_tasks):

    if task == other_task:

      task_results.append(1.)

      continue

    y_other = y_train[:, other_task]

    r2 = dc.metrics.pearson_r2_score(y_task, y_other)

    #print("r2 for %s-%s is %f" % (task, other_task, r2))

    task_results.append(r2)

  print("Task %d" % task)

  print(task_results)

  all_results.append(task_results)

print("Writing results to factors_corr.csv")

df = pd.DataFrame(all_results)

df.to_csv("factors_corr.csv")

    print("r2 for %s-%s is %f" % (task, other_task, r2))

    all_results.append(r2)

# the histogram of the data

n, bins, patches = plt.hist(np.array(all_results), 50, normed=True, stacked=True,

                            facecolor='green', alpha=0.75)

plt.xlabel('Cross-task Correlations')

plt.ylabel('Probability Density')

plt.title('Histogram of Factors Intertask Correlations')

plt.grid(True)

plt.savefig("Factors_correlations.png")

from FACTORS_datasets import load_factors

###Load data###

np.random.seed(123)

shard_size = 2000

num_trials = 5

print("About to load FACTORS data.")

import deepchem as dc

from FACTORS_datasets import load_factors

# Set numpy seed

np.random.seed(123)

###Load data###

shard_size = 2000

num_trials = 5

num_trials = 2

print("About to load FACTORS data.")

FACTORS_tasks, datasets, transformers = load_factors(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets

      weight_init_stddevs=[.02]*n_layers,

      bias_init_consts=[1.]*n_layers, learning_rate=.0003,

      penalty=.0001, penalty_type="l2", optimizer="adam", batch_size=100,

      seed=123)

      logdir="FACTORS_tf_model" )

  #Use R2 classification metric

  metric = dc.metrics.Metric(dc.metrics.pearson_r2_score, task_averager=np.mean)

import deepchem as dc

from FACTORS_datasets import load_factors

# Set numpy seed

np.random.seed(123)

###Load data###

shard_size = 2000

num_trials = 5

num_trials = 2

print("About to load FACTORS data.")

FACTORS_tasks, datasets, transformers = load_factors(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets

      alpha_init_stddevs=[.02]*n_layers, weight_init_stddevs=[.02]*n_layers,

      bias_init_consts=[1.]*n_layers, learning_rate=.0003,

      penalty=.0001, penalty_type="l2", optimizer="adam", batch_size=100,

      seed=123)

      logdir="FACTORS_tf_progressive")

  #Use R2 classification metric

  metric = dc.metrics.Metric(dc.metrics.pearson_r2_score, task_averager=np.mean)

import deepchem as dc

from FACTORS_datasets import load_factors

# Set numpy seed

np.random.seed(123)

###Load data###

shard_size = 2000

num_trials = 5

num_trials = 2

print("About to load FACTORS data.")

FACTORS_tasks, datasets, transformers = load_factors(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets

      bypass_weight_init_stddevs=[.02]*n_bypass_layers,

      bypass_bias_init_consts=[1.]*n_bypass_layers,

      learning_rate=.0003, penalty=.0001, penalty_type="l2",

      optimizer="adam", batch_size=100, seed=123)

      optimizer="adam", batch_size=100, logdir="FACTORS_tf_bypass")

  print("Fitting Model")

  model.fit(train_dataset, nb_epoch=nb_epoch)