feat(end-to-end): end-to-end prediction

biomes/predict.py

@@ -24,16 +24,18 @@ def predicted_map(B, change=0, path=None):
         ]
 
     frame = df[inputs + ['longitude']]
-    frame_cp = df[inputs + ['longitude']]
 
     for season in SEASONS:
         frame.loc[:, 'temp_{}_{}'.format(season, year)] += change
 
+    print(frame.head())
+    frame_cp = frame.copy()
+
     columns = ['latitude', 'longitude', 'biome_num']
     new_data = pd.DataFrame(columns=columns)
-    nframe = pd.DataFrame(columns=frame.columns, data=normalize_ndarray(frame.to_numpy(), frame_cp.to_numpy()))
+    nframe = pd.DataFrame(columns=frame.columns, data=normalize_ndarray(frame.to_numpy()))
 
-    for i, (chunk, chunk_original) in enumerate(zip(chunker(nframe, B.batch_size), chunker(frame, B.batch_size))):
+    for i, (chunk, chunk_original) in enumerate(zip(chunker(nframe, B.batch_size), chunker(frame_cp, B.batch_size))):
         if chunk.shape[0] < B.batch_size:
             continue
         input_data = chunk.loc[:, inputs].values
@@ -48,7 +50,7 @@ def predicted_map(B, change=0, path=None):
 
     draw(new_data, path=path)
 
-def predicted_map_cmd(checkpoint='checkpoints/save.h5', change=0, path=None):
+def predicted_map_cmd(checkpoint='checkpoints/b.h5', change=0, path=None):
     B = Model('b', epochs=1)
     B.prepare_for_use()
     B.restore(checkpoint)
@@ -136,6 +138,110 @@ def predicted_precips_cmd(checkpoint='checkpoints/precip.h5', year=2000):
     Precip.restore(checkpoint)
     predicted_precips(Precip, year=year)
 
-if __name__ == "__main__":
+def predict_end_to_end(Temp, Precip, Biomes, year=2000):
-    fire.Fire({ 'map': predicted_map_cmd, 'temp': predicted_temps_cmd, 'precip': predicted_precips_cmd })
+    columns = INPUTS
+
+    df = pd.read_pickle('data.p')
+
+    inputs = df[INPUTS]
+
+    all_temps = ['temp_{}_{}'.format(season, year) for season in SEASONS]
+    inputs.loc[:, 'mean_temp'] = np.mean(df[all_temps].values)
+
+    inputs = inputs.to_numpy()
+    inputs = normalize_ndarray(inputs)
+    out_columns = all_temps
+    out = Temp.predict(inputs)
+    temp_output = pd.DataFrame(data=denormalize(out, df[out_columns].to_numpy()), columns=out_columns)
+
+    inputs = df[INPUTS]
+
+    all_precips = ['precip_{}_{}'.format(season, year) for season in SEASONS]
+    inputs.loc[:, 'mean_precip'] = np.mean(df[all_precips].values)
+
+    inputs = inputs.to_numpy()
+    inputs = normalize_ndarray(inputs)
+    out_columns = all_precips
+    out = Precip.predict(inputs)
+
+    precip_output = pd.DataFrame(data=denormalize(out, df[out_columns].to_numpy()), columns=out_columns)
+
+    inputs = list(INPUTS)
+
+    for season in SEASONS:
+        inputs += [
+            'temp_{}_{}'.format(season, year),
+            'precip_{}_{}'.format(season, year)
+        ]
+
+    frame = df[inputs + ['longitude']]
+
+    for season in SEASONS:
+        tc = 'temp_{}_{}'.format(season, year)
+        pc = 'precip_{}_{}'.format(season, year)
+        frame.loc[:, tc] = temp_output[tc]
+        frame.loc[:, pc] = precip_output[pc]
+
+    frame.loc[:, 'latitude'] = df['latitude']
+
+    frame_cp = frame.copy()
+
+    columns = ['latitude', 'longitude', 'biome_num']
+    new_data = pd.DataFrame(columns=columns)
+    nframe = pd.DataFrame(columns=frame.columns, data=normalize_ndarray(frame.to_numpy()))
+
+    for i, (chunk, chunk_original) in enumerate(zip(chunker(nframe, Biomes.batch_size), chunker(frame_cp, Biomes.batch_size))):
+        if chunk.shape[0] < Biomes.batch_size:
+            continue
+        input_data = chunk.loc[:, inputs].values
+        out = Biomes.predict_class(input_data)
+
+        f = pd.DataFrame({
+            'longitude': chunk_original.loc[:, 'longitude'], 
+            'latitude': chunk_original.loc[:, 'latitude'],
+            'biome_num': out
+        }, columns=columns)
+        new_data = new_data.append(f)
+
+    print(new_data)
+    draw(new_data)
+
+def predict_end_to_end_cmd(checkpoint_temp='checkpoints/temp.h5', checkpoint_precip='checkpoints/precip.h5', checkpoint_biomes='checkpoints/b.h5', year=2000):
+    batch_size = A_params['batch_size']['grid_search'][0]
+    layers = A_params['layers']['grid_search'][0]
+    optimizer = A_params['optimizer']['grid_search'][0](A_params['lr']['grid_search'][0])
+
+    Temp = Model('temp', epochs=1)
+    Temp.prepare_for_use(
+        batch_size=batch_size,
+        layers=layers,
+        dataset_fn=dataframe_to_dataset_temp,
+        optimizer=optimizer,
+        out_activation=None,
+        loss='mse',
+        metrics=['mae']
+    )
+    Temp.restore(checkpoint_temp)
+
+    Precip = Model('precip', epochs=1)
+    Precip.prepare_for_use(
+        batch_size=batch_size,
+        layers=layers,
+        dataset_fn=dataframe_to_dataset_temp,
+        optimizer=optimizer,
+        out_activation=None,
+        loss='mse',
+        metrics=['mae']
+    )
+    Precip.restore(checkpoint_precip)
+
+    Biomes = Model('b', epochs=1)
+    Biomes.prepare_for_use()
+    Biomes.restore(checkpoint_biomes)
+
+    predict_end_to_end(Temp=Temp, Precip=Precip, Biomes=Biomes, year=year)
+
+
+if __name__ == "__main__":
+    fire.Fire({ 'map': predicted_map_cmd, 'temp': predicted_temps_cmd, 'precip': predicted_precips_cmd, 'end-to-end': predict_end_to_end_cmd })