Update app.py
Browse files
app.py
CHANGED
@@ -4,41 +4,69 @@ import numpy as np
|
|
4 |
import matplotlib.pyplot as plt
|
5 |
from io import StringIO
|
6 |
from momentfm import MOMENTPipeline
|
7 |
-
|
8 |
|
9 |
-
#
|
10 |
-
|
11 |
-
|
12 |
-
model_kwargs={"task_name": "anomaly_detection"}, # Changed task name
|
13 |
-
)
|
14 |
-
model.init()
|
15 |
|
16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
try:
|
18 |
-
|
19 |
-
|
|
|
|
|
20 |
|
21 |
-
# Validate
|
22 |
-
if
|
23 |
-
|
24 |
|
25 |
-
|
26 |
-
df =
|
|
|
|
|
|
|
|
|
|
|
27 |
|
28 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
values = df['value'].values.astype(np.float32)
|
30 |
-
values_3d = values.reshape(1, -1, 1) # Reshape to 3D
|
31 |
|
32 |
-
#
|
33 |
-
|
|
|
|
|
|
|
34 |
|
35 |
-
# Calculate
|
36 |
-
errors = np.abs(values -
|
37 |
|
38 |
-
# Dynamic threshold (
|
39 |
-
|
40 |
df['anomaly_score'] = errors
|
41 |
-
df['is_anomaly'] = errors >
|
42 |
|
43 |
# Create plot
|
44 |
fig, ax = plt.subplots(figsize=(12, 5))
|
@@ -48,30 +76,31 @@ def detect_anomalies(data_input, threshold=3.0): # Changed default threshold
|
|
48 |
df.loc[df['is_anomaly'], 'value'],
|
49 |
color='red', s=100, label='Anomaly'
|
50 |
)
|
51 |
-
ax.set_title(f'Anomaly Detection (Threshold: {
|
52 |
ax.legend()
|
53 |
|
54 |
# Prepare outputs
|
55 |
stats = {
|
56 |
"data_points": len(df),
|
57 |
-
"
|
58 |
-
"
|
59 |
-
"
|
60 |
}
|
61 |
|
62 |
return fig, stats, df.to_dict('records')
|
63 |
-
|
64 |
except Exception as e:
|
|
|
65 |
return None, {"error": str(e)}, None
|
66 |
|
67 |
-
# Gradio
|
68 |
-
with gr.Blocks() as demo:
|
69 |
-
gr.Markdown("
|
70 |
|
71 |
with gr.Row():
|
72 |
with gr.Column():
|
73 |
data_input = gr.Textbox(
|
74 |
-
label="Paste CSV
|
75 |
value="""timestamp,value
|
76 |
2025-04-01 00:00:00,100
|
77 |
2025-04-01 01:00:00,102
|
@@ -88,19 +117,28 @@ with gr.Blocks() as demo:
|
|
88 |
2025-04-01 12:00:00,101""",
|
89 |
lines=15
|
90 |
)
|
91 |
-
|
92 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
93 |
|
94 |
with gr.Column():
|
95 |
-
plot_output = gr.Plot()
|
96 |
-
stats_output = gr.JSON()
|
97 |
-
data_output = gr.JSON(
|
|
|
|
|
|
|
98 |
|
99 |
submit_btn.click(
|
100 |
detect_anomalies,
|
101 |
-
inputs=[data_input,
|
102 |
outputs=[plot_output, stats_output, data_output]
|
103 |
)
|
104 |
|
105 |
if __name__ == "__main__":
|
106 |
-
demo.launch()
|
|
|
4 |
import matplotlib.pyplot as plt
|
5 |
from io import StringIO
|
6 |
from momentfm import MOMENTPipeline
|
7 |
+
import logging
|
8 |
|
9 |
+
# Configure logging
|
10 |
+
logging.basicConfig(level=logging.INFO)
|
11 |
+
logger = logging.getLogger(__name__)
|
|
|
|
|
|
|
12 |
|
13 |
+
# Initialize model with reconstruction task
|
14 |
+
try:
|
15 |
+
model = MOMENTPipeline.from_pretrained(
|
16 |
+
"AutonLab/MOMENT-1-large",
|
17 |
+
model_kwargs={"task_name": "reconstruction"}, # Correct task name
|
18 |
+
)
|
19 |
+
model.init()
|
20 |
+
logger.info("Model loaded successfully")
|
21 |
+
except Exception as e:
|
22 |
+
logger.error(f"Model loading failed: {str(e)}")
|
23 |
+
raise
|
24 |
+
|
25 |
+
def validate_data(data_input):
|
26 |
+
"""Validate and process input data"""
|
27 |
try:
|
28 |
+
if isinstance(data_input, str):
|
29 |
+
df = pd.read_csv(StringIO(data_input))
|
30 |
+
else:
|
31 |
+
raise ValueError("Input must be CSV text")
|
32 |
|
33 |
+
# Validate columns
|
34 |
+
if not all(col in df.columns for col in ['timestamp', 'value']):
|
35 |
+
raise ValueError("CSV must contain 'timestamp' and 'value' columns")
|
36 |
|
37 |
+
# Convert timestamps
|
38 |
+
df['timestamp'] = pd.to_datetime(df['timestamp'], errors='coerce')
|
39 |
+
if df['timestamp'].isnull().any():
|
40 |
+
raise ValueError("Invalid timestamp format")
|
41 |
+
|
42 |
+
# Convert values to numeric
|
43 |
+
df['value'] = pd.to_numeric(df['value'], errors='raise')
|
44 |
|
45 |
+
return df.sort_values('timestamp')
|
46 |
+
|
47 |
+
except Exception as e:
|
48 |
+
logger.error(f"Data validation error: {str(e)}")
|
49 |
+
raise
|
50 |
+
|
51 |
+
def detect_anomalies(data_input, sensitivity=3.0):
|
52 |
+
"""Perform reconstruction-based anomaly detection"""
|
53 |
+
try:
|
54 |
+
df = validate_data(data_input)
|
55 |
values = df['value'].values.astype(np.float32)
|
|
|
56 |
|
57 |
+
# Reshape to 3D format (batch, sequence, features)
|
58 |
+
values_3d = values.reshape(1, -1, 1)
|
59 |
+
|
60 |
+
# Get reconstruction
|
61 |
+
reconstructed = model.reconstruct(values_3d)
|
62 |
|
63 |
+
# Calculate reconstruction error (MAE)
|
64 |
+
errors = np.abs(values - reconstructed[0,:,0])
|
65 |
|
66 |
+
# Dynamic threshold (z-score based)
|
67 |
+
threshold = np.mean(errors) + sensitivity * np.std(errors)
|
68 |
df['anomaly_score'] = errors
|
69 |
+
df['is_anomaly'] = errors > threshold
|
70 |
|
71 |
# Create plot
|
72 |
fig, ax = plt.subplots(figsize=(12, 5))
|
|
|
76 |
df.loc[df['is_anomaly'], 'value'],
|
77 |
color='red', s=100, label='Anomaly'
|
78 |
)
|
79 |
+
ax.set_title(f'Anomaly Detection (Threshold: {threshold:.2f})')
|
80 |
ax.legend()
|
81 |
|
82 |
# Prepare outputs
|
83 |
stats = {
|
84 |
"data_points": len(df),
|
85 |
+
"anomalous_points": int(df['is_anomaly'].sum()),
|
86 |
+
"detection_threshold": float(threshold),
|
87 |
+
"max_error": float(np.max(errors))
|
88 |
}
|
89 |
|
90 |
return fig, stats, df.to_dict('records')
|
91 |
+
|
92 |
except Exception as e:
|
93 |
+
logger.error(f"Detection error: {str(e)}")
|
94 |
return None, {"error": str(e)}, None
|
95 |
|
96 |
+
# Gradio Interface
|
97 |
+
with gr.Blocks(title="MOMENT Anomaly Detector") as demo:
|
98 |
+
gr.Markdown("## 🔍 Equipment Anomaly Detection using MOMENT")
|
99 |
|
100 |
with gr.Row():
|
101 |
with gr.Column():
|
102 |
data_input = gr.Textbox(
|
103 |
+
label="Paste time-series data (CSV format)",
|
104 |
value="""timestamp,value
|
105 |
2025-04-01 00:00:00,100
|
106 |
2025-04-01 01:00:00,102
|
|
|
117 |
2025-04-01 12:00:00,101""",
|
118 |
lines=15
|
119 |
)
|
120 |
+
sensitivity = gr.Slider(
|
121 |
+
minimum=1.0,
|
122 |
+
maximum=5.0,
|
123 |
+
value=3.0,
|
124 |
+
step=0.1,
|
125 |
+
label="Detection Sensitivity (Z-Score)"
|
126 |
+
)
|
127 |
+
submit_btn = gr.Button("Analyze Data", variant="primary")
|
128 |
|
129 |
with gr.Column():
|
130 |
+
plot_output = gr.Plot(label="Anomaly Detection Results")
|
131 |
+
stats_output = gr.JSON(label="Detection Statistics")
|
132 |
+
data_output = gr.JSON(
|
133 |
+
label="Processed Data",
|
134 |
+
max_lines=15
|
135 |
+
)
|
136 |
|
137 |
submit_btn.click(
|
138 |
detect_anomalies,
|
139 |
+
inputs=[data_input, sensitivity],
|
140 |
outputs=[plot_output, stats_output, data_output]
|
141 |
)
|
142 |
|
143 |
if __name__ == "__main__":
|
144 |
+
demo.launch(server_name="0.0.0.0", server_port=7860)
|