Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -15,8 +15,173 @@ def get_popular_tickers():
|
|
15 |
"JNJ", "V", "PG", "WMT", "BAC", "DIS", "NFLX", "INTC"
|
16 |
]
|
17 |
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
|
|
20 |
with gr.Blocks() as demo:
|
21 |
gr.Markdown("# Aplicaci贸n de Predicci贸n de Precios de Acciones")
|
22 |
|
@@ -24,47 +189,40 @@ with gr.Blocks() as demo:
|
|
24 |
with gr.Column(scale=1):
|
25 |
ticker = gr.Dropdown(
|
26 |
choices=get_popular_tickers(),
|
27 |
-
value="AAPL",
|
28 |
-
label="Selecciona el S铆mbolo de la Acci贸n"
|
29 |
-
|
30 |
-
train_data_points = gr.Slider(
|
31 |
-
minimum=50,
|
32 |
-
maximum=5000,
|
33 |
-
value=1000,
|
34 |
-
step=1,
|
35 |
-
label="N煤mero de Datos para Entrenamiento"
|
36 |
)
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
45 |
|
46 |
with gr.Column():
|
|
|
47 |
plot_output = gr.Plot(label="Gr谩fico de Predicci贸n")
|
48 |
download_btn = gr.File(label="Descargar Predicciones")
|
49 |
|
50 |
-
|
51 |
-
try:
|
52 |
-
stock = yf.Ticker(ticker)
|
53 |
-
hist = stock.history(period="max")
|
54 |
-
total_points = len(hist)
|
55 |
-
return gr.Slider.update(
|
56 |
-
maximum=total_points,
|
57 |
-
value=min(1000, total_points),
|
58 |
-
visible=True
|
59 |
-
)
|
60 |
-
except Exception as e:
|
61 |
-
print(f"Error updating slider: {str(e)}")
|
62 |
-
return gr.Slider.update(visible=True) # Mantener slider visible en caso de error
|
63 |
-
|
64 |
ticker.change(
|
65 |
fn=update_train_data_points,
|
66 |
inputs=[ticker],
|
67 |
-
outputs=[train_data_points]
|
|
|
68 |
)
|
69 |
|
70 |
predict_btn.click(
|
|
|
15 |
"JNJ", "V", "PG", "WMT", "BAC", "DIS", "NFLX", "INTC"
|
16 |
]
|
17 |
|
18 |
+
def predict_stock(ticker, train_data_points, prediction_days):
|
19 |
+
try:
|
20 |
+
# Asegurar que los par谩metros sean enteros
|
21 |
+
train_data_points = int(train_data_points)
|
22 |
+
prediction_days = int(prediction_days)
|
23 |
+
|
24 |
+
# Configurar el pipeline
|
25 |
+
pipeline = ChronosPipeline.from_pretrained(
|
26 |
+
"amazon/chronos-t5-mini",
|
27 |
+
device_map="cpu",
|
28 |
+
torch_dtype=torch.float32
|
29 |
+
)
|
30 |
+
|
31 |
+
# Obtener datos hist贸ricos
|
32 |
+
stock = yf.Ticker(ticker)
|
33 |
+
hist = stock.history(period="max")
|
34 |
+
if hist.empty:
|
35 |
+
raise ValueError(f"No hay datos disponibles para {ticker}")
|
36 |
+
|
37 |
+
stock_prices = hist[['Close']].reset_index()
|
38 |
+
df = stock_prices.rename(columns={'Date': 'Date', 'Close': f'{ticker}_Close'})
|
39 |
+
|
40 |
+
total_points = len(df)
|
41 |
+
if total_points < 50:
|
42 |
+
raise ValueError(f"Datos insuficientes para {ticker}")
|
43 |
+
|
44 |
+
# Asegurar que el n煤mero de datos de entrenamiento no exceda el total disponible
|
45 |
+
train_data_points = min(train_data_points, total_points)
|
46 |
+
|
47 |
+
# Crear el contexto para entrenamiento
|
48 |
+
context = torch.tensor(df[f'{ticker}_Close'][:train_data_points].values, dtype=torch.float32)
|
49 |
+
|
50 |
+
# Realizar predicci贸n
|
51 |
+
forecast = pipeline.predict(context, prediction_days, limit_prediction_length=False)
|
52 |
+
low, median, high = np.quantile(forecast[0].numpy(), [0.01, 0.5, 0.99], axis=0)
|
53 |
+
|
54 |
+
plt.figure(figsize=(20, 10))
|
55 |
+
plt.clf()
|
56 |
+
|
57 |
+
# Determinar el rango de fechas para mostrar
|
58 |
+
context_days = min(10, train_data_points)
|
59 |
+
start_index = max(0, train_data_points - context_days)
|
60 |
+
end_index = min(train_data_points + prediction_days, total_points)
|
61 |
+
|
62 |
+
# Plotear datos hist贸ricos
|
63 |
+
historical_dates = df['Date'][start_index:end_index]
|
64 |
+
historical_data = df[f'{ticker}_Close'][start_index:end_index].values
|
65 |
+
plt.plot(historical_dates,
|
66 |
+
historical_data,
|
67 |
+
color='blue',
|
68 |
+
linewidth=2,
|
69 |
+
label='Datos Reales')
|
70 |
+
|
71 |
+
# Crear fechas para la predicci贸n
|
72 |
+
if train_data_points < total_points:
|
73 |
+
prediction_start_date = df['Date'].iloc[train_data_points]
|
74 |
+
else:
|
75 |
+
last_date = df['Date'].iloc[-1]
|
76 |
+
prediction_start_date = last_date + pd.Timedelta(days=1)
|
77 |
+
|
78 |
+
prediction_dates = pd.date_range(start=prediction_start_date, periods=prediction_days, freq='B')
|
79 |
+
|
80 |
+
# Plotear predicci贸n
|
81 |
+
plt.plot(prediction_dates,
|
82 |
+
median,
|
83 |
+
color='black',
|
84 |
+
linewidth=2,
|
85 |
+
linestyle='-',
|
86 |
+
label='Predicci贸n')
|
87 |
+
|
88 |
+
# 脕rea de confianza
|
89 |
+
plt.fill_between(prediction_dates, low, high,
|
90 |
+
color='gray', alpha=0.2,
|
91 |
+
label='Intervalo de Confianza')
|
92 |
+
|
93 |
+
# Calcular m茅tricas si hay datos reales para comparar
|
94 |
+
overlap_end_index = train_data_points + prediction_days
|
95 |
+
if overlap_end_index <= total_points:
|
96 |
+
real_future_dates = df['Date'][train_data_points:overlap_end_index]
|
97 |
+
real_future_data = df[f'{ticker}_Close'][train_data_points:overlap_end_index].values
|
98 |
+
|
99 |
+
matching_dates = real_future_dates[real_future_dates.isin(prediction_dates)]
|
100 |
+
matching_indices = matching_dates.index - train_data_points
|
101 |
+
plt.plot(matching_dates,
|
102 |
+
real_future_data[matching_indices],
|
103 |
+
color='red',
|
104 |
+
linewidth=2,
|
105 |
+
linestyle='--',
|
106 |
+
label='Datos Reales de Validaci贸n')
|
107 |
+
|
108 |
+
predicted_data = median[:len(matching_indices)]
|
109 |
+
mae = mean_absolute_error(real_future_data[matching_indices], predicted_data)
|
110 |
+
rmse = np.sqrt(mean_squared_error(real_future_data[matching_indices], predicted_data))
|
111 |
+
mape = np.mean(np.abs((real_future_data[matching_indices] - predicted_data) / real_future_data[matching_indices])) * 100
|
112 |
+
plt.title(f"Predicci贸n del Precio de {ticker}\nMAE: {mae:.2f} | RMSE: {rmse:.2f} | MAPE: {mape:.2f}%",
|
113 |
+
fontsize=14, pad=20)
|
114 |
+
else:
|
115 |
+
plt.title(f"Predicci贸n Futura del Precio de {ticker}",
|
116 |
+
fontsize=14, pad=20)
|
117 |
+
|
118 |
+
plt.legend(loc="upper left", fontsize=12)
|
119 |
+
plt.xlabel("Fecha", fontsize=12)
|
120 |
+
plt.ylabel("Precio", fontsize=12)
|
121 |
+
|
122 |
+
plt.grid(True, which='both', axis='x', linestyle='--', linewidth=0.5)
|
123 |
+
|
124 |
+
ax = plt.gca()
|
125 |
+
locator = mdates.DayLocator()
|
126 |
+
formatter = mdates.DateFormatter('%Y-%m-%d')
|
127 |
+
ax.xaxis.set_major_locator(locator)
|
128 |
+
ax.xaxis.set_major_formatter(formatter)
|
129 |
+
|
130 |
+
plt.setp(ax.get_xticklabels(), rotation=45, ha='right')
|
131 |
+
|
132 |
+
plt.tight_layout()
|
133 |
+
|
134 |
+
# Crear archivo CSV temporal
|
135 |
+
temp_csv = tempfile.NamedTemporaryFile(delete=False, suffix='.csv')
|
136 |
+
prediction_df = pd.DataFrame({
|
137 |
+
'Date': prediction_dates,
|
138 |
+
'Predicted_Price': median,
|
139 |
+
'Lower_Bound': low,
|
140 |
+
'Upper_Bound': high
|
141 |
+
})
|
142 |
+
|
143 |
+
if overlap_end_index <= total_points:
|
144 |
+
real_future_dates = df['Date'][train_data_points:overlap_end_index]
|
145 |
+
real_future_data = df[f'{ticker}_Close'][train_data_points:overlap_end_index].values
|
146 |
+
matching_dates = real_future_dates[real_future_dates.isin(prediction_dates)]
|
147 |
+
prediction_df = prediction_df[prediction_df['Date'].isin(matching_dates)]
|
148 |
+
prediction_df['Real_Price'] = real_future_data[:len(prediction_df)]
|
149 |
+
|
150 |
+
prediction_df.to_csv(temp_csv.name, index=False)
|
151 |
+
temp_csv.close()
|
152 |
+
|
153 |
+
return plt, temp_csv.name
|
154 |
+
|
155 |
+
except Exception as e:
|
156 |
+
print(f"Error: {str(e)}")
|
157 |
+
raise gr.Error(f"Error al procesar {ticker}: {str(e)}")
|
158 |
+
|
159 |
+
def update_train_data_points(ticker):
|
160 |
+
if not ticker:
|
161 |
+
return gr.Slider.update(value=1000, maximum=5000)
|
162 |
+
|
163 |
+
try:
|
164 |
+
stock = yf.Ticker(ticker)
|
165 |
+
hist = stock.history(period="max")
|
166 |
+
if hist.empty:
|
167 |
+
raise ValueError(f"No hay datos disponibles para {ticker}")
|
168 |
+
|
169 |
+
total_points = len(hist)
|
170 |
+
if total_points < 50:
|
171 |
+
raise ValueError(f"Datos insuficientes para {ticker}")
|
172 |
+
|
173 |
+
return gr.Slider.update(
|
174 |
+
maximum=total_points,
|
175 |
+
value=min(1000, total_points),
|
176 |
+
minimum=50,
|
177 |
+
step=1,
|
178 |
+
interactive=True
|
179 |
+
)
|
180 |
+
except Exception as e:
|
181 |
+
print(f"Error al actualizar datos para {ticker}: {str(e)}")
|
182 |
+
return gr.Slider.update(value=1000, maximum=5000, minimum=50, step=1)
|
183 |
|
184 |
+
# Interfaz de Gradio
|
185 |
with gr.Blocks() as demo:
|
186 |
gr.Markdown("# Aplicaci贸n de Predicci贸n de Precios de Acciones")
|
187 |
|
|
|
189 |
with gr.Column(scale=1):
|
190 |
ticker = gr.Dropdown(
|
191 |
choices=get_popular_tickers(),
|
192 |
+
value="AAPL",
|
193 |
+
label="Selecciona el S铆mbolo de la Acci贸n",
|
194 |
+
interactive=True
|
|
|
|
|
|
|
|
|
|
|
|
|
195 |
)
|
196 |
+
with gr.Column():
|
197 |
+
train_data_points = gr.Slider(
|
198 |
+
minimum=50,
|
199 |
+
maximum=5000,
|
200 |
+
value=1000,
|
201 |
+
step=1,
|
202 |
+
label="N煤mero de Datos para Entrenamiento",
|
203 |
+
interactive=True
|
204 |
+
)
|
205 |
+
prediction_days = gr.Slider(
|
206 |
+
minimum=1,
|
207 |
+
maximum=60,
|
208 |
+
value=5,
|
209 |
+
step=1,
|
210 |
+
label="N煤mero de D铆as a Predecir",
|
211 |
+
interactive=True
|
212 |
+
)
|
213 |
+
predict_btn = gr.Button("Predecir", interactive=True)
|
214 |
|
215 |
with gr.Column():
|
216 |
+
error_output = gr.Textbox(label="Estado", visible=False)
|
217 |
plot_output = gr.Plot(label="Gr谩fico de Predicci贸n")
|
218 |
download_btn = gr.File(label="Descargar Predicciones")
|
219 |
|
220 |
+
# Eventos
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
221 |
ticker.change(
|
222 |
fn=update_train_data_points,
|
223 |
inputs=[ticker],
|
224 |
+
outputs=[train_data_points],
|
225 |
+
api_name="update_data"
|
226 |
)
|
227 |
|
228 |
predict_btn.click(
|