updates

.gitignore

@@ -1,3 +1,5 @@
+maps
+logs
 checkpoints.*
 geodata
 *.p

draw.py

@@ -1,10 +1,11 @@
 from shapely.geometry import Point, MultiPoint
 from shapely.ops import cascaded_union
 import matplotlib.pyplot as plt
+
 import pandas as pd
 import cartopy.crs as ccrs
 
-def draw(df):
+def draw(df, path=None):
     biomes = {}
     biome_numbers = df['biome_num'].unique()
     # biome_names = df['biome_name'].unique()
@@ -48,6 +49,9 @@ def draw(df):
         # print(artist.get_label())
 
     # ax.legend(artists, biome_names)
+    if path:
+        plt.savefig(path)
+    else:
         plt.show()
 
 if __name__ == "__main__":

nn.py

@@ -22,7 +22,7 @@ np.random.seed(RANDOM_SEED)
 df = pd.read_pickle('data_final.p')
 
 class Model():
-    def __init__(self, name, batch_size=100, shuffle_buffer_size=500, learning_rate=0.001, epochs=1):
+    def __init__(self, name, batch_size=64, shuffle_buffer_size=500, learning_rate=0.001, epochs=1):
         self.name = name
         self.path = "checkpoints/{}.hdf5".format(name)
 
@@ -55,13 +55,13 @@ class Model():
             keras.layers.Dense(self.output_size)
         ])
 
-    def compile(self):
+    def compile(self, loss='mse', metrics=['accuracy'], optimizer=tf.train.AdamOptimizer):
-        self.model.load_weights(self.path)
+        # self.model.load_weights(self.path)
-        optimizer = tf.train.AdamOptimizer(self.learning_rate)
+        optimizer = optimizer(self.learning_rate)
 
-        self.model.compile(loss='mse',
+        self.model.compile(loss=loss,
                     optimizer=optimizer,
-                    metrics=['mae', 'accuracy'])
+                    metrics=metrics)
 
     def evaluate(self):
         return self.model.evaluate(
@@ -74,39 +74,45 @@ class Model():
     def train(self):
         self.model.summary()
 
-        checkpoint = keras.callbacks.ModelCheckpoint(filepath, monitor='acc', verbose=1, mode='max')
+        checkpoint = keras.callbacks.ModelCheckpoint(self.path, monitor='acc', verbose=1, mode='max')
+        tensorboard = keras.callbacks.TensorBoard(log_dir='./logs')
+        # map_callback = keras.callbacks.LambdaCallback(on_epoch_end=self.map_callback)
 
         self.model.fit(
             self.training,
             batch_size=self.batch_size,
             epochs=self.epochs,
             steps_per_epoch=int(self.dataset_size / self.batch_size),
-            callbacks=[checkpoint],
+            callbacks=[checkpoint, tensorboard],
             verbose=1
         )
 
     def predict(self, a):
         return np.argmax(self.model.predict(a), axis=1)
 
-A = Model('a', batch_size=100, shuffle_buffer_size=500, learning_rate=0.001, epochs=2)
+A = Model('a', epochs=2)
-B = Model('b', batch_size=100, shuffle_buffer_size=500, learning_rate=0.001, epochs=850)
+B = Model('b', learning_rate=0.001, batch_size=100, epochs=100)
 
-if __name__ == "__main__":
+def compile_b():
     B.prepare_dataset(df, dataframe_to_dataset_biomes)
     B.create_model([64, 128])
-    B.compile()
+    B.compile(loss='sparse_categorical_crossentropy')
+
+def compile_a():
+    A.prepare_dataset(df, dataframe_to_dataset_temp_precip)
+    A.create_model([4])
+    A.compile(metrics=['accuracy', 'mae'])
+
+if __name__ == "__main__":
+    compile_b()
+    B.train()
 
     # for inp, out in B.test.take(1).make_one_shot_iterator():
         # print(inp, out)
 
     # print(np.unique(nums))
     # print(np.unique(predictions))
+    # print('loss: {}, evaluation: {}'.format(*B.evaluate()))
 
-    print('loss: {}, evaluation: {}'.format(*B.evaluate()))
+    compile_a()
-
+    A.train()
-    # B.train()
-
-    A.prepare_dataset(df, dataframe_to_dataset_temp_precip)
-    A.create_model([4])
-    A.compile()
-    # A.train()

predict.py

@@ -14,11 +14,9 @@ df = pd.read_pickle('data_final.p')
 latitude = np.array(df.index.get_level_values(1))
 df.loc[:, 'latitude'] = pd.Series(latitude, index=df.index)
 
-B.prepare_dataset(df, dataframe_to_dataset_biomes)
+compile_b()
-B.create_model([64, 128])
-B.compile()
 
-for change in range(-5, 6):
+for change in range(0, 1):
     print('TEMPERATURE MODIFICATION OF {}'.format(change))
 
     inputs = ['elevation', 'distance_to_water']
@@ -34,15 +32,8 @@ for change in range(-5, 6):
     frame = df[inputs]
     print(frame.head())
 
-    # for season in SEASONS:
+    for season in SEASONS:
-        # frame.loc[:, 'temp_{}_{}'.format(season, year)] += change
+        frame.loc[:, 'temp_{}_{}'.format(season, year)] += change
-
-    # print(np.average(frame.loc[:, 'temp_winter_2016']))
-
-    # index = []
-    # for longitude in range(-179, 179):
-        # for latitude in range(-89, 89):
-            # index.append((longitude, latitude))
 
     columns = ['biome_num']
     new_data = pd.DataFrame(columns=columns)
@@ -54,12 +45,4 @@ for change in range(-5, 6):
         new_data = new_data.reindex(new_index)
         new_data.loc[chunk.index.values, 'biome_num'] = out
 
-    # print(new_data['biome_num'].unique())
-
     draw(new_data)
-
-    # columns = ['biome_num']
-    # indices = ['longitude', 'latitude']
-    # new_df = pd.DataFrame(index=indices, columns=columns)
-    # new_df = 
-