notebook.community

Edit and run

Comparing Accuracy: scikit-learn vs tensorflow

In this notebook, we train then test the model in a 60 / 40 split for the decision tree algo on both scikit-learn and tensorflow. First, we start with scikit-learn where we predict cloud vendor based on throughput.





In [1]:



    


import graphviz
import pandas
from sklearn import tree
from sklearn.model_selection import train_test_split

clf = tree.DecisionTreeClassifier()
input = pandas.read_csv("/home/glenn/git/clojure-news-feed/client/ml/etl/throughput.csv")
data = input[input.columns[6:9]]
target = input['cloud']
X_train, X_test, y_train, y_test = train_test_split(data, target, test_size=0.4, random_state=0)
clf = clf.fit(X_train, y_train)
clf.score(X_test, y_test)



    


















    
Out[1]:








0.937007874015748

Next, we evaluate scikit-learn accuracy where we predict feed implementation based on latency.





In [1]:



    


import graphviz
import pandas
from sklearn import tree
from sklearn.model_selection import train_test_split

clf = tree.DecisionTreeClassifier()
input = pandas.read_csv("/home/glenn/git/clojure-news-feed/client/ml/etl/latency.csv")
data = input[input.columns[7:9]]
data['cloud'] = input['cloud'].apply(lambda x: 1.0 if x == 'GKE' else 0.0)
target = input['feed']
X_train, X_test, y_train, y_test = train_test_split(data, target, test_size=0.4, random_state=0)
clf = clf.fit(X_train, y_train)
clf.score(X_test, y_test)



    


















    






/home/glenn/.local/lib/python2.7/site-packages/ipykernel_launcher.py:9: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy
  if __name__ == '__main__':










    
Out[1]:








0.9874015748031496

As you can see, scikit-learn has a 99% accuracy rate. We now do the same thing with tensorflow.





In [1]:



    


import tensorflow as tf
import numpy as np
import pandas
from tensorflow.python.ops import parsing_ops
from tensorflow.contrib.tensor_forest.python import tensor_forest
from tensorflow.contrib.learn.python.learn.utils import input_fn_utils
from sklearn.model_selection import train_test_split

input = pandas.read_csv("/home/glenn/git/clojure-news-feed/client/ml/etl/latency.csv")
data = input[input.columns[7:9]]
data['cloud'] = input['cloud'].apply(lambda x: 1.0 if x == 'GKE' else 0.0)
X_train, X_test, y_train, y_test = train_test_split(data, input['feed'], test_size=0.4, random_state=0)
X_train_np = np.array(X_train, dtype=np.float32)
y_train_np = np.array(y_train, dtype=np.int32)
X_test_np = np.array(X_test, dtype=np.float32)
y_test_np = np.array(y_test, dtype=np.int32)
hparams = tensor_forest.ForestHParams(num_classes=7,
                                      num_features=3,
                                      num_trees=1,
                                      regression=False,
                                      max_nodes=500).fill()
classifier = tf.contrib.tensor_forest.client.random_forest.TensorForestEstimator(hparams)
c = classifier.fit(x=X_train_np, y=y_train_np)
c.evaluate(x=X_test_np, y=y_test_np)



    


















    






WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/tensor_forest/client/random_forest.py:121: multi_class_head (from tensorflow.contrib.learn.python.learn.estimators.head) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.contrib.estimator.*_head.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:1179: __init__ (from tensorflow.contrib.learn.python.learn.estimators.estimator) is deprecated and will be removed in a future version.
Instructions for updating:
Please replace uses of any Estimator from tf.contrib.learn with an Estimator from tf.estimator.*
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:427: __init__ (from tensorflow.contrib.learn.python.learn.estimators.run_config) is deprecated and will be removed in a future version.
Instructions for updating:
When switching to tf.estimator.Estimator, use tf.estimator.RunConfig instead.
INFO:tensorflow:Using default config.
WARNING:tensorflow:Using temporary folder as model directory: /tmp/tmpFG57Uw
INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_num_ps_replicas': 0, '_keep_checkpoint_max': 5, '_task_type': None, '_train_distribute': None, '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7f5a8068a4d0>, '_model_dir': '/tmp/tmpFG57Uw', '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_session_config': None, '_tf_random_seed': None, '_save_summary_steps': 100, '_device_fn': None, '_num_worker_replicas': 0, '_task_id': 0, '_log_step_count_steps': 100, '_tf_config': gpu_options {
  per_process_gpu_memory_fraction: 1.0
}
, '_evaluation_master': '', '_environment': 'local', '_master': ''}
WARNING:tensorflow:From <ipython-input-1-1482161d90f6>:23: calling fit (from tensorflow.contrib.learn.python.learn.estimators.estimator) with y is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
WARNING:tensorflow:From <ipython-input-1-1482161d90f6>:23: calling fit (from tensorflow.contrib.learn.python.learn.estimators.estimator) with x is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))










    






/home/glenn/.local/lib/python2.7/site-packages/ipykernel_launcher.py:11: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy
  # This is added back by InteractiveShellApp.init_path()










    






WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:508: __init__ (from tensorflow.contrib.learn.python.learn.estimators.estimator) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to the Estimator interface.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:142: setup_train_data_feeder (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please use tensorflow/transform or tf.data.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:100: extract_pandas_data (from tensorflow.contrib.learn.python.learn.learn_io.pandas_io) is deprecated and will be removed in a future version.
Instructions for updating:
Please access pandas data directly.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:102: extract_pandas_labels (from tensorflow.contrib.learn.python.learn.learn_io.pandas_io) is deprecated and will be removed in a future version.
Instructions for updating:
Please access pandas data directly.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:159: __init__ (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please use tensorflow/transform or tf.data.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:340: check_array (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please convert numpy dtypes explicitly.
INFO:tensorflow:Constructing forest with params = 
INFO:tensorflow:{'num_output_columns': 8, 'feature_bagging_fraction': 1.0, 'valid_leaf_threshold': 1, 'checkpoint_stats': False, 'initialize_average_splits': False, 'pruning_type': 0, 'prune_every_samples': 0, 'dominate_fraction': 0.99, 'max_fertile_nodes': 0, 'early_finish_check_every_samples': 0, 'dominate_method': 'bootstrap', 'bagging_fraction': 1.0, 'regression': False, 'param_file': None, 'bagged_num_features': 3, 'use_running_stats_method': False, 'max_nodes': 500, 'split_finish_name': 'basic', 'leaf_model_type': 0, 'stats_model_type': 0, 'bagged_features': None, 'num_features': 3, 'split_after_samples': 250, 'num_outputs': 1, 'collate_examples': False, 'split_type': 0, 'num_classes': 7, 'num_splits_to_consider': 10, 'split_name': 'less_or_equal', 'finish_type': 0, 'inference_tree_paths': False, 'split_pruning_name': 'none', 'base_random_seed': 0, 'num_trees': 1, 'model_name': 'all_dense'}
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/head.py:678: __new__ (from tensorflow.contrib.learn.python.learn.estimators.model_fn) is deprecated and will be removed in a future version.
Instructions for updating:
When switching to tf.estimator.Estimator, use tf.estimator.EstimatorSpec. You can use the `estimator_spec` method to create an equivalent one.
INFO:tensorflow:Create CheckpointSaverHook.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:Saving checkpoints for 0 into /tmp/tmpFG57Uw/model.ckpt.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:loss = 1.9459084, step = 1
INFO:tensorflow:TensorForestLossHook resetting last_step.
INFO:tensorflow:global_step/sec: 460.562
INFO:tensorflow:loss = 1.1720499, step = 101 (0.219 sec)
INFO:tensorflow:global_step/sec: 489.096
INFO:tensorflow:loss = 1.1666828, step = 201 (0.206 sec)
INFO:tensorflow:global_step/sec: 454.219
INFO:tensorflow:loss = 1.165421, step = 301 (0.220 sec)
INFO:tensorflow:TensorForestLossHook requesting stop.
INFO:tensorflow:Saving checkpoints for 317 into /tmp/tmpFG57Uw/model.ckpt.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:9: clean up resources: None
INFO:tensorflow:Loss for final step: 1.165421.
WARNING:tensorflow:From <ipython-input-1-1482161d90f6>:24: calling evaluate (from tensorflow.contrib.learn.python.learn.estimators.estimator) with y is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
WARNING:tensorflow:From <ipython-input-1-1482161d90f6>:24: calling evaluate (from tensorflow.contrib.learn.python.learn.estimators.estimator) with x is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
INFO:tensorflow:Constructing forest with params = 
INFO:tensorflow:{'num_output_columns': 8, 'params_proto': pruning_type {
  prune_every_samples {
    constant_value: 0.0
  }
}
finish_type {
  check_every_steps {
    constant_value: 0.0
  }
}
num_trees: 1
max_nodes: 500
num_outputs: 7
num_splits_to_consider {
  constant_value: 10.0
}
split_after_samples {
  constant_value: 250.0
}
dominate_fraction {
  constant_value: 0.990000009537
}
num_features: 3
, 'feature_bagging_fraction': 1.0, 'valid_leaf_threshold': 1, 'checkpoint_stats': False, 'initialize_average_splits': False, 'pruning_type': 0, 'prune_every_samples': 0, 'dominate_fraction': 0.99, 'max_fertile_nodes': 0, 'early_finish_check_every_samples': 0, 'dominate_method': 'bootstrap', 'bagging_fraction': 1.0, 'regression': False, 'param_file': None, 'bagged_num_features': 3, 'use_running_stats_method': False, 'max_nodes': 500, 'split_finish_name': 'basic', 'leaf_model_type': 0, 'stats_model_type': 0, 'bagged_features': None, 'num_features': 3, 'split_after_samples': 250, 'num_outputs': 1, 'serialized_params_proto': '"\x07\n\x05\r\x00\x00\x00\x00*\x07\n\x05\r\x00\x00\x00\x000\x018\xf4\x03`\x07j\x05\r\x00\x00 Ar\x05\r\x00\x00zCz\x05\r\xa4p}?\xa8\x01\x03', 'collate_examples': False, 'split_type': 0, 'num_classes': 7, 'num_splits_to_consider': 10, 'split_name': 'less_or_equal', 'finish_type': 0, 'inference_tree_paths': False, 'split_pruning_name': 'none', 'base_random_seed': 0, 'num_trees': 1, 'model_name': 'all_dense'}
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
INFO:tensorflow:Starting evaluation at 2018-11-08-07:13:54
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/tmpFG57Uw/model.ckpt-317
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Finished evaluation at 2018-11-08-07:13:54
INFO:tensorflow:Saving dict for global step 317: accuracy = 0.97952753, global_step = 317, loss = 1.1858935
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'










    
Out[1]:








{'accuracy': 0.97952753, 'global_step': 317, 'loss': 1.1858935}

Looks like tensorflow has a 98% accuracy rate which is 1% less than scikit-learn algo. Let us use Tensorflow to look at the accuracy of predicting cloud vendor based on throughput.





In [1]:



    


import tensorflow as tf
import numpy as np
import pandas
from tensorflow.python.ops import parsing_ops
from tensorflow.contrib.tensor_forest.python import tensor_forest
from tensorflow.contrib.learn.python.learn.utils import input_fn_utils
from sklearn.model_selection import train_test_split

input = pandas.read_csv("/home/glenn/git/clojure-news-feed/client/ml/etl/throughput.csv")
data = input[input.columns[6:9]]
target = input['cloud'].apply(lambda x: 1.0 if x == 'GKE' else 0.0)
X_train, X_test, y_train, y_test = train_test_split(data, target, test_size=0.4, random_state=0)
X_train_np = np.array(X_train, dtype=np.float32)
y_train_np = np.array(y_train, dtype=np.int32)
X_test_np = np.array(X_test, dtype=np.float32)
y_test_np = np.array(y_test, dtype=np.int32)
hparams = tensor_forest.ForestHParams(num_classes=3,
                                      num_features=3,
                                      num_trees=1,
                                      regression=False,
                                      max_nodes=500).fill()
classifier = tf.contrib.tensor_forest.client.random_forest.TensorForestEstimator(hparams)
c = classifier.fit(x=X_train_np, y=y_train_np)
c.evaluate(x=X_test_np, y=y_test_np)



    


















    






WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/tensor_forest/client/random_forest.py:121: multi_class_head (from tensorflow.contrib.learn.python.learn.estimators.head) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.contrib.estimator.*_head.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:1179: __init__ (from tensorflow.contrib.learn.python.learn.estimators.estimator) is deprecated and will be removed in a future version.
Instructions for updating:
Please replace uses of any Estimator from tf.contrib.learn with an Estimator from tf.estimator.*
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:427: __init__ (from tensorflow.contrib.learn.python.learn.estimators.run_config) is deprecated and will be removed in a future version.
Instructions for updating:
When switching to tf.estimator.Estimator, use tf.estimator.RunConfig instead.
INFO:tensorflow:Using default config.
WARNING:tensorflow:Using temporary folder as model directory: /tmp/tmpdLcZIL
INFO:tensorflow:Using config: {'_save_checkpoints_secs': 600, '_num_ps_replicas': 0, '_keep_checkpoint_max': 5, '_task_type': None, '_train_distribute': None, '_is_chief': True, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x7ff632e00fd0>, '_model_dir': '/tmp/tmpdLcZIL', '_save_checkpoints_steps': None, '_keep_checkpoint_every_n_hours': 10000, '_session_config': None, '_tf_random_seed': None, '_save_summary_steps': 100, '_device_fn': None, '_num_worker_replicas': 0, '_task_id': 0, '_log_step_count_steps': 100, '_tf_config': gpu_options {
  per_process_gpu_memory_fraction: 1.0
}
, '_evaluation_master': '', '_environment': 'local', '_master': ''}
WARNING:tensorflow:From <ipython-input-1-a0403f9ccc2d>:23: calling fit (from tensorflow.contrib.learn.python.learn.estimators.estimator) with y is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
WARNING:tensorflow:From <ipython-input-1-a0403f9ccc2d>:23: calling fit (from tensorflow.contrib.learn.python.learn.estimators.estimator) with x is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:508: __init__ (from tensorflow.contrib.learn.python.learn.estimators.estimator) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to the Estimator interface.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:142: setup_train_data_feeder (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please use tensorflow/transform or tf.data.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:100: extract_pandas_data (from tensorflow.contrib.learn.python.learn.learn_io.pandas_io) is deprecated and will be removed in a future version.
Instructions for updating:
Please access pandas data directly.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:102: extract_pandas_labels (from tensorflow.contrib.learn.python.learn.learn_io.pandas_io) is deprecated and will be removed in a future version.
Instructions for updating:
Please access pandas data directly.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:159: __init__ (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please use tensorflow/transform or tf.data.
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py:340: check_array (from tensorflow.contrib.learn.python.learn.learn_io.data_feeder) is deprecated and will be removed in a future version.
Instructions for updating:
Please convert numpy dtypes explicitly.
INFO:tensorflow:Constructing forest with params = 
INFO:tensorflow:{'num_output_columns': 4, 'feature_bagging_fraction': 1.0, 'valid_leaf_threshold': 1, 'checkpoint_stats': False, 'initialize_average_splits': False, 'pruning_type': 0, 'prune_every_samples': 0, 'dominate_fraction': 0.99, 'max_fertile_nodes': 0, 'early_finish_check_every_samples': 0, 'dominate_method': 'bootstrap', 'bagging_fraction': 1.0, 'regression': False, 'param_file': None, 'bagged_num_features': 3, 'use_running_stats_method': False, 'max_nodes': 500, 'split_finish_name': 'basic', 'leaf_model_type': 0, 'stats_model_type': 0, 'bagged_features': None, 'num_features': 3, 'split_after_samples': 250, 'num_outputs': 1, 'collate_examples': False, 'split_type': 0, 'num_classes': 3, 'num_splits_to_consider': 10, 'split_name': 'less_or_equal', 'finish_type': 0, 'inference_tree_paths': False, 'split_pruning_name': 'none', 'base_random_seed': 0, 'num_trees': 1, 'model_name': 'all_dense'}
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
WARNING:tensorflow:From /home/glenn/.local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/head.py:678: __new__ (from tensorflow.contrib.learn.python.learn.estimators.model_fn) is deprecated and will be removed in a future version.
Instructions for updating:
When switching to tf.estimator.Estimator, use tf.estimator.EstimatorSpec. You can use the `estimator_spec` method to create an equivalent one.
INFO:tensorflow:Create CheckpointSaverHook.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:Saving checkpoints for 0 into /tmp/tmpdLcZIL/model.ckpt.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:loss = 1.0986115, step = 1
INFO:tensorflow:TensorForestLossHook resetting last_step.
INFO:tensorflow:global_step/sec: 435.284
INFO:tensorflow:loss = 0.5794245, step = 101 (0.232 sec)
INFO:tensorflow:global_step/sec: 502.091
INFO:tensorflow:loss = 0.5637042, step = 201 (0.199 sec)
INFO:tensorflow:global_step/sec: 487.722
INFO:tensorflow:loss = 0.5540101, step = 301 (0.207 sec)
INFO:tensorflow:global_step/sec: 323.398
INFO:tensorflow:loss = 0.5523, step = 401 (0.307 sec)
INFO:tensorflow:TensorForestLossHook requesting stop.
INFO:tensorflow:Saving checkpoints for 459 into /tmp/tmpdLcZIL/model.ckpt.
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'
INFO:tensorflow:9: clean up resources: None
INFO:tensorflow:Loss for final step: 0.5522999.
WARNING:tensorflow:From <ipython-input-1-a0403f9ccc2d>:24: calling evaluate (from tensorflow.contrib.learn.python.learn.estimators.estimator) with y is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
WARNING:tensorflow:From <ipython-input-1-a0403f9ccc2d>:24: calling evaluate (from tensorflow.contrib.learn.python.learn.estimators.estimator) with x is deprecated and will be removed after 2016-12-01.
Instructions for updating:
Estimator is decoupled from Scikit Learn interface by moving into
separate class SKCompat. Arguments x, y and batch_size are only
available in the SKCompat class, Estimator will only accept input_fn.
Example conversion:
  est = Estimator(...) -> est = SKCompat(Estimator(...))
INFO:tensorflow:Constructing forest with params = 
INFO:tensorflow:{'num_output_columns': 4, 'params_proto': pruning_type {
  prune_every_samples {
    constant_value: 0.0
  }
}
finish_type {
  check_every_steps {
    constant_value: 0.0
  }
}
num_trees: 1
max_nodes: 500
num_outputs: 3
num_splits_to_consider {
  constant_value: 10.0
}
split_after_samples {
  constant_value: 250.0
}
dominate_fraction {
  constant_value: 0.990000009537
}
num_features: 3
, 'feature_bagging_fraction': 1.0, 'valid_leaf_threshold': 1, 'checkpoint_stats': False, 'initialize_average_splits': False, 'pruning_type': 0, 'prune_every_samples': 0, 'dominate_fraction': 0.99, 'max_fertile_nodes': 0, 'early_finish_check_every_samples': 0, 'dominate_method': 'bootstrap', 'bagging_fraction': 1.0, 'regression': False, 'param_file': None, 'bagged_num_features': 3, 'use_running_stats_method': False, 'max_nodes': 500, 'split_finish_name': 'basic', 'leaf_model_type': 0, 'stats_model_type': 0, 'bagged_features': None, 'num_features': 3, 'split_after_samples': 250, 'num_outputs': 1, 'serialized_params_proto': '"\x07\n\x05\r\x00\x00\x00\x00*\x07\n\x05\r\x00\x00\x00\x000\x018\xf4\x03`\x03j\x05\r\x00\x00 Ar\x05\r\x00\x00zCz\x05\r\xa4p}?\xa8\x01\x03', 'collate_examples': False, 'split_type': 0, 'num_classes': 3, 'num_splits_to_consider': 10, 'split_name': 'less_or_equal', 'finish_type': 0, 'inference_tree_paths': False, 'split_pruning_name': 'none', 'base_random_seed': 0, 'num_trees': 1, 'model_name': 'all_dense'}
INFO:tensorflow:dense_features_size: 3 dense: [{name: features original_type: 0 size: 3}] sparse: []
INFO:tensorflow:Starting evaluation at 2018-11-22-17:17:23
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Restoring parameters from /tmp/tmpdLcZIL/model.ckpt-459
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
INFO:tensorflow:Finished evaluation at 2018-11-22-17:17:24
INFO:tensorflow:Saving dict for global step 459: accuracy = 0.9149606, global_step = 459, loss = 0.6355508
WARNING:tensorflow:Issue encountered when serializing resources.
Type is unsupported, or the types of the items don't match field type in CollectionDef. Note this is a warning and probably safe to ignore.
'_Resource' object has no attribute 'name'










    
Out[1]:








{'accuracy': 0.9149606, 'global_step': 459, 'loss': 0.6355508}

Looks like Tensorflow is 3% less accurate than Scikit-learn when it comes to predicting cloud vendor based on throughput.





In [ ]:

Content source: gengstrand/clojure-news-feed

Similar notebooks:

notebook.community | gallery | about