image

app.py

@@ -16,6 +16,8 @@ from pytorch_forecasting import (
     TemporalFusionTransformer,
 )
 
+from PIL import Image
+
 ## Functions 
 def raw_preds_to_df(raw,quantiles = None):
     """
@@ -38,7 +40,6 @@ def raw_preds_to_df(raw,quantiles = None):
     preds_df['pred_idx'] = preds_df['time_idx'] + preds_df['h'] - 1
     return preds_df
 
-#@st.cache_data
 def prepare_dataset(_parameters, df, rain, temperature, datepicker, mapping):
     if rain != "Default":
         df["MTXWTH_Day_precip"] = mapping[rain]
@@ -54,13 +55,11 @@ def prepare_dataset(_parameters, df, rain, temperature, datepicker, mapping):
     df = TimeSeriesDataSet.from_parameters(_parameters, df)
     return df.to_dataloader(train=False, batch_size=256,num_workers = 0)
 
-#@st.cache_data
 def predict(_model, _dataloader, datepicker):
     out = _model.predict(_dataloader, mode="raw", return_x=True, return_index=True)
     preds = raw_preds_to_df(out, quantiles = None)
     return preds[["pred_idx", "Group", "pred"]]
 
-#@st.cache_data
 def adjust_data_for_plot(df, preds):
     df = pd.merge(df, preds, left_on=["time_idx", "Group"], right_on=["pred_idx", "Group"], how = "left")
     df = df[~df["pred"].isna()]
@@ -111,7 +110,11 @@ def init_model():
 def main():
     # Start App
     st.title("Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting")
-    
+
+
+    image = Image.open('data/image.png')
+
+    st.image(image, caption='Coding.Waterkant Festival for AI')
     st.markdown(body = """
         ### Abstract
         Multi-horizon forecasting often contains a complex mix of inputs – including
@@ -132,6 +135,10 @@ def main():
         interpretability use cases of TFT.
     
         ### Experiments
+        We implemented TFT for sales multi-horizon sales forecast during Coding.Waterkant. 
+        Please try our implementation and adjust some of the training data.
+
+        Adjustments to the model and extention with Quantile forecasts are coming soon ;) 
     """)
     
     try:
@@ -158,12 +165,14 @@ def main():
 
     data_plot = adjust_data_for_plot(df.copy(), preds)
     fig, _ = generate_plot(data_plot)
-    
-    st.pyplot(fig)
-    rain = st.radio("Rain Indicator", ('Default', 'Yes', 'No'), key = "rain")
-    temperature = st.slider('Change in Temperature', min_value=-10.0, max_value=10.0, value=0.0, step=0.25, key = "temperature")
+
     datepicker = st.date_input("Start of Forecast", datetime.date(2022, 10, 24), min_value=datetime.date(2022, 6, 26) + datetime.timedelta(days = 35), max_value=datetime.date(2023, 6, 26) - datetime.timedelta(days = 30), key = "date")
 
+    st.pyplot(fig)
+
+    temperature = st.slider('Change in Temperature', min_value=-10.0, max_value=10.0, value=0.0, step=0.25, key = "temperature")
+    rain = st.radio("Rain Indicator", ('Default', 'Yes', 'No'), key = "rain")
+    
 if __name__ == '__main__':
     main()