Make Training Great Again!

pywatts/__init__.py

@@ -1,3 +1,4 @@
 from pywatts import db
 from pywatts import fetchdata
 from pywatts import neural
+from pywatts import main

pywatts/main.py

@@ -0,0 +1,13 @@
+import tensorflow as tf
+import pywatts.neural
+
+from sklearn.model_selection import train_test_split
+
+df = pywatts.db.rows_to_df(list(range(1, 50)))
+X = df[[col for col in df.columns if col != 'dc']]
+y = df['dc']
+
+X_train, X_tmp, y_train, y_tmp = train_test_split(X, y, test_size=0.2, random_state=23)
+
+feature_cols = [tf.feature_column.numeric_column(col) for col in X.columns]
+n = pywatts.neural.Net(feature_cols=feature_cols)

pywatts/neural.py

@@ -1,5 +1,6 @@
 import tensorflow as tf
 
+
 def pywatts_input_fn(X, y=None, num_epochs=None, shuffle=True, batch_size=400):
     return tf.estimator.inputs.pandas_input_fn(x=X,
                                                y=y,
@@ -12,14 +13,13 @@ class Net:
     __regressor = None
     __feature_cols = [tf.feature_column.numeric_column(col) for col in ['dc', 'temp', 'wind']]
 
-
     def __init__(self, feature_cols=__feature_cols):
         self.__regressor = tf.estimator.DNNRegressor(feature_columns=feature_cols,
                                                      hidden_units=[50, 50],
                                                      model_dir='tf_pywatts_model')
 
-    def train(self, training_data, steps):
-        self.__regressor.train(input_fn=pywatts_input_fn(training_data, num_epochs=None, shuffle=True), steps=steps)
+    def train(self, training_data, training_results, steps):
+        self.__regressor.train(input_fn=pywatts_input_fn(training_data, y=training_results, num_epochs=None, shuffle=True), steps=steps)
 
     def evaluate(self, eval_data):
         self.__regressor.evaluate(input_fn=self.pywatts_input_fn(eval_data, num_epochs=1, shuffle=False), steps=1)