import tensorflow as tf
import os
# Just disables the warning, doesn't enable AVX/FMA
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import numpy as np
import pandas as pd
census = pd.read_csv("/Users/rajacsp/datasets/census_data.csv")
|
age |
workclass |
education |
education_num |
marital_status |
occupation |
relationship |
race |
gender |
capital_gain |
capital_loss |
hours_per_week |
native_country |
income_bracket |
| 0 |
39 |
State-gov |
Bachelors |
13 |
Never-married |
Adm-clerical |
Not-in-family |
White |
Male |
2174 |
0 |
40 |
United-States |
<=50K |
| 1 |
50 |
Self-emp-not-inc |
Bachelors |
13 |
Married-civ-spouse |
Exec-managerial |
Husband |
White |
Male |
0 |
0 |
13 |
United-States |
<=50K |
| 2 |
38 |
Private |
HS-grad |
9 |
Divorced |
Handlers-cleaners |
Not-in-family |
White |
Male |
0 |
0 |
40 |
United-States |
<=50K |
| 3 |
53 |
Private |
11th |
7 |
Married-civ-spouse |
Handlers-cleaners |
Husband |
Black |
Male |
0 |
0 |
40 |
United-States |
<=50K |
| 4 |
28 |
Private |
Bachelors |
13 |
Married-civ-spouse |
Prof-specialty |
Wife |
Black |
Female |
0 |
0 |
40 |
Cuba |
<=50K |
census['income_bracket'].unique()
array([' <=50K', ' >50K'], dtype=object)
# Fix the income bracket
def label_fix(label):
label = label.strip().lower()
if(label == '<=50k'):
return 0
return 1
census['income_bracket'] = census['income_bracket'].apply(label_fix)
census['income_bracket'].unique()
array([0, 1])
# Train Test Split
from sklearn.model_selection import train_test_split
x_data = census.drop('income_bracket', axis=1)
y_labels = census['income_bracket']
X_train, x_test, y_train, y_est = train_test_split(x_data, y_labels, test_size=0.3, random_state=101)
Index(['age', 'workclass', 'education', 'education_num', 'marital_status',
'occupation', 'relationship', 'race', 'gender', 'capital_gain',
'capital_loss', 'hours_per_week', 'native_country', 'income_bracket'],
dtype='object')
import tensorflow as tf
from tensorflow.feature_column import categorical_column_with_vocabulary_list as vlist
gender = vlist("gender", ["Female", "Male"])
VocabularyListCategoricalColumn(key='gender', vocabulary_list=('Female', 'Male'), dtype=tf.string, default_value=-1, num_oov_buckets=0)
from tensorflow.feature_column import categorical_column_with_hash_bucket as bucket
occupation = bucket("occupation", hash_bucket_size=1000)
marital_status = bucket("marital_status", hash_bucket_size=1000)
relationship = bucket("relationship", hash_bucket_size=1000)
education = bucket("education", hash_bucket_size=1000)
workclass = bucket("workclass", hash_bucket_size=1000)
native_country = bucket("native_country", hash_bucket_size=1000)
from tensorflow.feature_column import numeric_column as nc
age = nc("age")
education_num = nc("education_num")
capital_gain = nc("capital_gain")
capital_loss = nc("capital_loss")
hours_per_week = nc("hours_per_week")
feat_cols = [gender, occupation, marital_status, relationship, education, workclass, native_country, age, education_num,
capital_gain, capital_loss, hours_per_week]
[VocabularyListCategoricalColumn(key='gender', vocabulary_list=('Female', 'Male'), dtype=tf.string, default_value=-1, num_oov_buckets=0),
HashedCategoricalColumn(key='occupation', hash_bucket_size=1000, dtype=tf.string),
HashedCategoricalColumn(key='marital_status', hash_bucket_size=1000, dtype=tf.string),
HashedCategoricalColumn(key='relationship', hash_bucket_size=1000, dtype=tf.string),
HashedCategoricalColumn(key='education', hash_bucket_size=1000, dtype=tf.string),
HashedCategoricalColumn(key='workclass', hash_bucket_size=1000, dtype=tf.string),
HashedCategoricalColumn(key='native_country', hash_bucket_size=1000, dtype=tf.string),
NumericColumn(key='age', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None),
NumericColumn(key='education_num', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None),
NumericColumn(key='capital_gain', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None),
NumericColumn(key='capital_loss', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None),
NumericColumn(key='hours_per_week', shape=(1,), default_value=None, dtype=tf.float32, normalizer_fn=None)]
input_func = tf.estimator.inputs.pandas_input_fn(x=X_train, y = y_train, batch_size=100, num_epochs=None, shuffle=True)
model = tf.estimator.LinearClassifier(feature_columns=feat_cols)
INFO:tensorflow:Using default config.
WARNING:tensorflow:Using temporary folder as model directory: /var/folders/fk/b_m_3t9915zc2llvfkwd4d_h0000gn/T/tmp3xkmlm9h
INFO:tensorflow:Using config: {'_model_dir': '/var/folders/fk/b_m_3t9915zc2llvfkwd4d_h0000gn/T/tmp3xkmlm9h', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': None, '_save_checkpoints_secs': 600, '_session_config': allow_soft_placement: true
graph_options {
rewrite_options {
meta_optimizer_iterations: ONE
}
}
, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_service': None, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x130e60e80>, '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}
model.train(input_fn=input_func, steps=5000)
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.
Instructions for updating:
Colocations handled automatically by placer.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow_estimator/python/estimator/inputs/queues/feeding_queue_runner.py:62: QueueRunner.__init__ (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.
Instructions for updating:
To construct input pipelines, use the `tf.data` module.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow_estimator/python/estimator/inputs/queues/feeding_functions.py:500: add_queue_runner (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.
Instructions for updating:
To construct input pipelines, use the `tf.data` module.
INFO:tensorflow:Calling model_fn.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow/python/feature_column/feature_column_v2.py:2703: to_float (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow/python/ops/lookup_ops.py:1137: to_int64 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.
Instructions for updating:
Use tf.cast instead.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Create CheckpointSaverHook.
INFO:tensorflow:Graph was finalized.
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow/python/training/monitored_session.py:809: start_queue_runners (from tensorflow.python.training.queue_runner_impl) is deprecated and will be removed in a future version.
Instructions for updating:
To construct input pipelines, use the `tf.data` module.
INFO:tensorflow:Saving checkpoints for 0 into /var/folders/fk/b_m_3t9915zc2llvfkwd4d_h0000gn/T/tmp3xkmlm9h/model.ckpt.
INFO:tensorflow:loss = 69.31472, step = 1
INFO:tensorflow:global_step/sec: 183.162
INFO:tensorflow:loss = 74.34316, step = 101 (0.549 sec)
INFO:tensorflow:global_step/sec: 282.272
INFO:tensorflow:loss = 258.81635, step = 201 (0.354 sec)
INFO:tensorflow:global_step/sec: 360.103
INFO:tensorflow:loss = 92.93664, step = 301 (0.277 sec)
INFO:tensorflow:global_step/sec: 310.323
INFO:tensorflow:loss = 479.21606, step = 401 (0.323 sec)
INFO:tensorflow:global_step/sec: 363.822
INFO:tensorflow:loss = 64.496826, step = 501 (0.273 sec)
INFO:tensorflow:global_step/sec: 287.337
INFO:tensorflow:loss = 142.80907, step = 601 (0.349 sec)
INFO:tensorflow:global_step/sec: 329.981
INFO:tensorflow:loss = 158.61514, step = 701 (0.303 sec)
INFO:tensorflow:global_step/sec: 361.742
INFO:tensorflow:loss = 92.889305, step = 801 (0.276 sec)
INFO:tensorflow:global_step/sec: 336.053
INFO:tensorflow:loss = 47.939163, step = 901 (0.298 sec)
INFO:tensorflow:global_step/sec: 386.539
INFO:tensorflow:loss = 109.61961, step = 1001 (0.258 sec)
INFO:tensorflow:global_step/sec: 349.966
INFO:tensorflow:loss = 97.20078, step = 1101 (0.287 sec)
INFO:tensorflow:global_step/sec: 367.692
INFO:tensorflow:loss = 102.36524, step = 1201 (0.272 sec)
INFO:tensorflow:global_step/sec: 386.157
INFO:tensorflow:loss = 97.24974, step = 1301 (0.258 sec)
INFO:tensorflow:global_step/sec: 426.666
INFO:tensorflow:loss = 106.33708, step = 1401 (0.234 sec)
INFO:tensorflow:global_step/sec: 390.189
INFO:tensorflow:loss = 88.20786, step = 1501 (0.258 sec)
INFO:tensorflow:global_step/sec: 392.219
INFO:tensorflow:loss = 40.560345, step = 1601 (0.254 sec)
INFO:tensorflow:global_step/sec: 408.792
INFO:tensorflow:loss = 67.237366, step = 1701 (0.243 sec)
INFO:tensorflow:global_step/sec: 392.664
INFO:tensorflow:loss = 100.42717, step = 1801 (0.257 sec)
INFO:tensorflow:global_step/sec: 391.489
INFO:tensorflow:loss = 104.48391, step = 1901 (0.254 sec)
INFO:tensorflow:global_step/sec: 346.702
INFO:tensorflow:loss = 30.307816, step = 2001 (0.289 sec)
INFO:tensorflow:global_step/sec: 381.499
INFO:tensorflow:loss = 41.16053, step = 2101 (0.261 sec)
INFO:tensorflow:global_step/sec: 292.709
INFO:tensorflow:loss = 73.87081, step = 2201 (0.342 sec)
INFO:tensorflow:global_step/sec: 341.69
INFO:tensorflow:loss = 42.253674, step = 2301 (0.295 sec)
INFO:tensorflow:global_step/sec: 333.577
INFO:tensorflow:loss = 32.43901, step = 2401 (0.300 sec)
INFO:tensorflow:global_step/sec: 380.409
INFO:tensorflow:loss = 37.70379, step = 2501 (0.262 sec)
INFO:tensorflow:global_step/sec: 340.751
INFO:tensorflow:loss = 39.260693, step = 2601 (0.294 sec)
INFO:tensorflow:global_step/sec: 291.395
INFO:tensorflow:loss = 51.416866, step = 2701 (0.343 sec)
INFO:tensorflow:global_step/sec: 315.53
INFO:tensorflow:loss = 184.38156, step = 2801 (0.316 sec)
INFO:tensorflow:global_step/sec: 366.986
INFO:tensorflow:loss = 53.05883, step = 2901 (0.272 sec)
INFO:tensorflow:global_step/sec: 265.167
INFO:tensorflow:loss = 27.692787, step = 3001 (0.380 sec)
INFO:tensorflow:global_step/sec: 339.134
INFO:tensorflow:loss = 39.770794, step = 3101 (0.292 sec)
INFO:tensorflow:global_step/sec: 358.485
INFO:tensorflow:loss = 47.001534, step = 3201 (0.279 sec)
INFO:tensorflow:global_step/sec: 281.97
INFO:tensorflow:loss = 36.61994, step = 3301 (0.356 sec)
INFO:tensorflow:global_step/sec: 343.422
INFO:tensorflow:loss = 46.674873, step = 3401 (0.289 sec)
INFO:tensorflow:global_step/sec: 281.989
INFO:tensorflow:loss = 36.499893, step = 3501 (0.356 sec)
INFO:tensorflow:global_step/sec: 352.627
INFO:tensorflow:loss = 44.82026, step = 3601 (0.282 sec)
INFO:tensorflow:global_step/sec: 357.209
INFO:tensorflow:loss = 109.4581, step = 3701 (0.280 sec)
INFO:tensorflow:global_step/sec: 289.537
INFO:tensorflow:loss = 52.670895, step = 3801 (0.349 sec)
INFO:tensorflow:global_step/sec: 337.938
INFO:tensorflow:loss = 25.518808, step = 3901 (0.294 sec)
INFO:tensorflow:global_step/sec: 326.968
INFO:tensorflow:loss = 88.96236, step = 4001 (0.306 sec)
INFO:tensorflow:global_step/sec: 335.375
INFO:tensorflow:loss = 34.12282, step = 4101 (0.296 sec)
INFO:tensorflow:global_step/sec: 330.084
INFO:tensorflow:loss = 26.477894, step = 4201 (0.304 sec)
INFO:tensorflow:global_step/sec: 342.657
INFO:tensorflow:loss = 58.477943, step = 4301 (0.292 sec)
INFO:tensorflow:global_step/sec: 320.761
INFO:tensorflow:loss = 109.7404, step = 4401 (0.313 sec)
INFO:tensorflow:global_step/sec: 322.263
INFO:tensorflow:loss = 70.80423, step = 4501 (0.309 sec)
INFO:tensorflow:global_step/sec: 276.567
INFO:tensorflow:loss = 42.813023, step = 4601 (0.362 sec)
INFO:tensorflow:global_step/sec: 335.875
INFO:tensorflow:loss = 50.28676, step = 4701 (0.296 sec)
INFO:tensorflow:global_step/sec: 263.377
INFO:tensorflow:loss = 69.291405, step = 4801 (0.382 sec)
INFO:tensorflow:global_step/sec: 297.762
INFO:tensorflow:loss = 112.38106, step = 4901 (0.336 sec)
INFO:tensorflow:Saving checkpoints for 5000 into /var/folders/fk/b_m_3t9915zc2llvfkwd4d_h0000gn/T/tmp3xkmlm9h/model.ckpt.
INFO:tensorflow:Loss for final step: 32.800762.
<tensorflow_estimator.python.estimator.canned.linear.LinearClassifier at 0x130e60940>
# Evaluation
pred_fn = tf.estimator.inputs.pandas_input_fn(x=x_test, batch_size=len(x_test), shuffle=False)
prediction = list(model.predict(input_fn=pred_fn))
INFO:tensorflow:Calling model_fn.
INFO:tensorflow:Done calling model_fn.
INFO:tensorflow:Graph was finalized.
WARNING:tensorflow:From /Users/rajacsp/anaconda3/envs/py36/lib/python3.6/site-packages/tensorflow/python/training/saver.py:1266: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.
Instructions for updating:
Use standard file APIs to check for files with this prefix.
INFO:tensorflow:Restoring parameters from /var/folders/fk/b_m_3t9915zc2llvfkwd4d_h0000gn/T/tmp3xkmlm9h/model.ckpt-5000
INFO:tensorflow:Running local_init_op.
INFO:tensorflow:Done running local_init_op.
{'class_ids': array([0]),
'classes': array([b'0'], dtype=object),
'logistic': array([0.24731174], dtype=float32),
'logits': array([-1.1130015], dtype=float32),
'probabilities': array([0.7526882 , 0.24731174], dtype=float32)}
final_preds = []
for pred in prediction:
final_preds.append(pred['class_ids'][0])
[0, 0, 0, 0, 1, 0, 0, 0, 0, 0]
from sklearn.metrics import classification_report
print(classification_report(y_est, final_preds))
precision recall f1-score support
0 0.86 0.93 0.89 7436
1 0.70 0.54 0.61 2333
micro avg 0.83 0.83 0.83 9769
macro avg 0.78 0.73 0.75 9769
weighted avg 0.83 0.83 0.83 9769