Commit inicial.

app.py

@@ -0,0 +1,83 @@
+from kats.models.prophet import ProphetModel, ProphetParams
+from kats.consts import TimeSeriesData
+import pandas as pd
+import matplotlib.pyplot as plt
+import gradio as gr
+
+def build_model(time_series_csv, changepoint_range, changepoint_prior_scale, seasonality_prior_scale, seasonality_mode, test_size):
+  
+  # time_series_csv = pd.DataFrame(time_series_csv)
+  time_series_csv = pd.read_csv(time_series_csv.name, delimiter=",")
+
+  # Divisón entre entrenamiento y test (sintaxis de Kats)
+  train_size = 1 - test_size
+  split_idx = int(len(time_series_csv) * train_size)
+  
+  ts_train = TimeSeriesData(
+            pd.DataFrame(
+                {
+                    "time": time_series_csv['time'][:split_idx],
+                    "target": time_series_csv['value'][:split_idx]
+                }
+            )
+        )
+  
+  ts_test = TimeSeriesData(
+            pd.DataFrame(
+                {
+                    "time": time_series_csv['time'][split_idx:],
+                    "target": time_series_csv['value'][split_idx:]
+                }
+            )
+        )
+
+
+  # Creación del modelo
+  prophet_params = ProphetParams(
+        changepoint_range = changepoint_range,
+        changepoint_prior_scale = changepoint_prior_scale,
+        seasonality_prior_scale = seasonality_prior_scale,
+        seasonality_mode = seasonality_mode
+    )
+  
+  model = ProphetModel(ts_train, prophet_params)
+    
+  # Entrenamiento del modelo
+  model.fit()
+
+  # Predicción
+  forecasting = model.predict(
+        steps=len(ts_test)
+    )
+  
+  result = pd.concat([
+            ts_test.to_dataframe().set_index('time'),
+            forecasting.set_index('time')
+          ], axis=1)
+
+  fig = plt.figure()
+  plt.plot(result)
+  plt.title("Prediction")
+
+  return fig
+
+
+# Interfaz
+iface = gr.Interface(
+    fn=build_model, 
+    inputs=[
+            gr.File(),
+            gr.Slider(0.8, 0.99),
+            gr.Number(precision=2),
+            gr.Number(precision=2),
+            gr.Radio(choices=['additive','multiplicative']),
+            gr.Slider(0.1,0.35,step=0.1,value=0.2)
+    ], 
+    outputs=gr.Plot(),
+    title="Prophet model generator",
+    description="Upload your .csv with a time series and set up your own Prophet model to make predictions. 🔮",
+    allow_flagging='never',
+    theme="peach"
+)
+
+iface.launch(debug=True)

requirements.txt

@@ -0,0 +1,3 @@
+kats
+pandas
+matplotlib