Commit a0d83754 authored by Bharath Ramsundar's avatar Bharath Ramsundar
Browse files

UV changes

parent e17f245a

examples/uv/UV_correlations.py

0 → 100644
+43 −0
Original line number Diff line number Diff line 
"""

Script that computes correlations of UV tasks. 

"""

from __future__ import print_function

from __future__ import division

from __future__ import unicode_literals



import os

import numpy as np

import tempfile

import shutil

import deepchem as dc

import pandas as pd

from UV_datasets import load_uv



###Load data###

np.random.seed(123)

shard_size = 2000

print("About to load UV data.")

UV_tasks, datasets, transformers = load_uv(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets



y_train = train_dataset.y

n_tasks = y_train.shape[1]



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 uv_corr.csv")

df = pd.DataFrame(all_results)

df.to_csv("uv_corr.csv")

examples/uv/UV_tf_model.py

+71 −31
Original line number Diff line number Diff line
@@ -11,19 +11,25 @@ import tempfile 
import shutil

import numpy as np

import deepchem as dc

from MERCK_datasets import load_uv

from UV_datasets import load_uv



# Set numpy seed

np.random.seed(123)



###Load data###

shard_size = 2000

num_shards_per_batch = 4

print("About to load MERCK data.")

UV_tasks, datasets, transformers = load_uv(

    shard_size=shard_size, num_shards_per_batch=num_shards_per_batch)

num_trials = 1

print("About to load UV data.")

UV_tasks, datasets, transformers = load_uv(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets



#############################################################  DEBUG

print("np.amin(train_dataset.y)")

print(np.amin(train_dataset.y))

print("np.amax(train_dataset.y)")

print(np.amax(train_dataset.y))

#############################################################  DEBUG



print("Number of compounds in train set")

print(len(train_dataset))

print("Number of compounds in validation set")
@@ -31,10 +37,10 @@ print(len(valid_dataset)) 
print("Number of compounds in test set")

print(len(test_dataset))



all_results = []

for trial in range(num_trials):

  ###Create model###

#n_layers = 3

n_layers = 2

#nb_epoch = 30 

  n_layers = 3

  nb_epoch = 50

  model = dc.models.TensorflowMultiTaskRegressor(

      len(UV_tasks), train_dataset.get_data_shape()[0],
@@ -42,7 +48,7 @@ model = dc.models.TensorflowMultiTaskRegressor( 
      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, verbosity="high")

      seed=123, logdir="UV_tf_model")



  #Use R2 classification metric

  metric = dc.metrics.Metric(dc.metrics.pearson_r2_score, task_averager=np.mean)
@@ -50,16 +56,50 @@ metric = dc.metrics.Metric(dc.metrics.pearson_r2_score, task_averager=np.mean) 
  print("Training model")

  model.fit(train_dataset, nb_epoch=nb_epoch)



train_scores = model.evaluate(train_dataset, [metric], transformers)

valid_scores = model.evaluate(valid_dataset, [metric], transformers)

#Only use for final evaluation

test_scores = model.evaluate(test_dataset, [metric], transformers)

  print("Evaluating models")

  train_score, train_task_scores = model.evaluate(

      train_dataset, [metric], transformers, per_task_metrics=True)

  valid_score, valid_task_scores = model.evaluate(

      valid_dataset, [metric], transformers, per_task_metrics=True)

  test_score, test_task_scores = model.evaluate(

      test_dataset, [metric], transformers, per_task_metrics=True)



  all_results.append((train_score, train_task_scores,

                      valid_score, valid_task_scores,

                      test_score, test_task_scores))



  print("Scores for trial %d" % trial)

  print("----------------------------------------------------------------")

  print("train_task_scores")

  print(train_task_scores)

  print("Mean Train score")

  print(train_score)

  print("valid_task_scores")

  print(valid_task_scores)

  print("Mean Validation score")

  print(valid_score)

  print("test_task_scores")

  print(test_task_scores)

  print("Mean Test score")

  print(test_score)



print("Train scores")

print(train_scores)

print("####################################################################")



print("Validation scores")

print(valid_scores)

for trial in range(num_trials):

  (train_score, train_task_scores, valid_score, valid_task_scores,

   test_score, test_task_scores) = all_results[trial]



print("Test scores")

print(test_scores)
 
  print("Scores for trial %d" % trial)

  print("----------------------------------------------------------------")

  print("train_task_scores")

  print(train_task_scores)

  print("Mean Train score")

  print(train_score)

  print("valid_task_scores")

  print(valid_task_scores)

  print("Mean Validation score")

  print(valid_score)

  print("test_task_scores")

  print(test_task_scores)

  print("Mean Test score")

  print(test_score)

examples/uv/UV_tf_robust.py

+63 −26
Original line number Diff line number Diff line
@@ -11,17 +11,17 @@ import numpy as np 
import tempfile

import shutil

import deepchem as dc

from MERCK_datasets import load_uv

from UV_datasets import load_uv



# Set numpy seed

np.random.seed(123)



###Load data###

shard_size = 2000

num_shards_per_batch = 4

print("About to load MERCK data.")

UV_tasks, datasets, transformers = load_uv(

    shard_size=shard_size, num_shards_per_batch=num_shards_per_batch)

#num_trials = 5

num_trials = 1

print("About to load data.")

UV_tasks, datasets, transformers = load_uv(shard_size=shard_size)

train_dataset, valid_dataset, test_dataset = datasets



print("Number of compounds in train set")
@@ -34,6 +34,12 @@ print(len(test_dataset)) 
n_layers = 3

n_bypass_layers = 3

nb_epoch = 30



#Use R2 classification metric

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



all_results = []

for trial in range(num_trials):

  model = dc.models.RobustMultitaskRegressor(

      len(UV_tasks), train_dataset.get_data_shape()[0],

      layer_sizes=[500]*n_layers, bypass_layer_sizes=[40]*n_bypass_layers,
@@ -42,24 +48,55 @@ model = dc.models.RobustMultitaskRegressor( 
      bypass_weight_init_stddevs=[.02]*n_bypass_layers,

      bypass_bias_init_consts=[.5]*n_bypass_layers,

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

    optimizer="adam", batch_size=100, verbosity="high")

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



#Use R2 classification metric

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



print("Optimizing Hyperparameters")

  print("Fitting Model")

  model.fit(train_dataset, nb_epoch=nb_epoch)



train_scores = model.evaluate(train_dataset, [metric], transformers)

valid_scores = model.evaluate(valid_dataset, [metric], transformers)

#Only use for final evaluation

test_scores = model.evaluate(test_dataset, [metric], transformers)

  print("Evaluating models")

  train_score, train_task_scores = model.evaluate(

      train_dataset, [metric], transformers, per_task_metrics=True)

  valid_score, valid_task_scores = model.evaluate(

      valid_dataset, [metric], transformers, per_task_metrics=True)

  test_score, test_task_scores = model.evaluate(

      test_dataset, [metric], transformers, per_task_metrics=True)



  all_results.append((train_score, train_task_scores,

                      valid_score, valid_task_scores,

                      test_score, test_task_scores))



  print("Scores for trial %d" % trial)

  print("----------------------------------------------------------------")

  print("train_task_scores")

  print(train_task_scores)

  print("Mean Train score")

  print(train_score)

  print("valid_task_scores")

  print(valid_task_scores)

  print("Mean Validation score")

  print(valid_score)

  print("test_task_scores")

  print(test_task_scores)

  print("Mean Test score")

  print(test_score)



print("Train scores")

print(train_scores)

print("####################################################################")



print("Validation scores")

print(valid_scores)

for trial in range(num_trials):

  (train_score, train_task_scores, valid_score, valid_task_scores,

   test_score, test_task_scores) = all_results[trial]



print("Test scores")

print(test_scores)
 
  print("Scores for trial %d" % trial)

  print("----------------------------------------------------------------")

  print("train_task_scores")

  print(train_task_scores)

  print("Mean Train score")

  print(train_score)

  print("valid_task_scores")

  print(valid_task_scores)

  print("Mean Validation score")

  print(valid_score)

  print("test_task_scores")

  print(test_task_scores)

  print("Mean Test score")

  print(test_score)

CRA Git | Maintained and supported by SUSTech CRA and CCSE