Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -2,11 +2,11 @@ import gradio as gr
|
|
2 |
import numpy as np
|
3 |
import pandas as pd
|
4 |
from sklearn.cluster import KMeans
|
5 |
-
from sklearn.metrics import pairwise_distances_argmin_min
|
6 |
-
import seaborn as sns
|
7 |
import matplotlib.pyplot as plt
|
|
|
8 |
import io
|
9 |
-
import os
|
10 |
from PIL import Image
|
11 |
|
12 |
# Define the paths for example data
|
@@ -23,10 +23,10 @@ EXAMPLE_FILES = {
|
|
23 |
|
24 |
class Clusters:
|
25 |
def __init__(self, loc_vars):
|
26 |
-
self.kmeans = KMeans(n_clusters=1000, random_state=0, n_init=10).fit(np.ascontiguousarray(loc_vars))
|
27 |
-
closest, _ = pairwise_distances_argmin_min(
|
28 |
|
29 |
-
rep_ids = pd.Series(data=(closest+1))
|
30 |
rep_ids.name = 'policy_id'
|
31 |
rep_ids.index.name = 'cluster_id'
|
32 |
self.rep_ids = rep_ids
|
@@ -34,6 +34,7 @@ class Clusters:
|
|
34 |
self.policy_count = self.agg_by_cluster(pd.DataFrame({'policy_count': [1] * len(loc_vars)}))['policy_count']
|
35 |
|
36 |
def agg_by_cluster(self, df, agg=None):
|
|
|
37 |
temp = df.copy()
|
38 |
temp['cluster_id'] = self.kmeans.labels_
|
39 |
temp = temp.set_index('cluster_id')
|
@@ -41,14 +42,17 @@ class Clusters:
|
|
41 |
return temp.groupby(temp.index).agg(agg)
|
42 |
|
43 |
def extract_reps(self, df):
|
|
|
44 |
temp = pd.merge(self.rep_ids, df.reset_index(), how='left', on='policy_id')
|
45 |
temp.index.name = 'cluster_id'
|
46 |
return temp.drop('policy_id', axis=1)
|
47 |
|
48 |
def extract_and_scale_reps(self, df, agg=None):
|
|
|
49 |
if agg:
|
50 |
cols = df.columns
|
51 |
mult = pd.DataFrame({c: (self.policy_count if (c not in agg or agg[c] == 'sum') else 1) for c in cols})
|
|
|
52 |
extracted_df = self.extract_reps(df)
|
53 |
mult.index = extracted_df.index
|
54 |
return extracted_df.mul(mult)
|
@@ -56,47 +60,57 @@ class Clusters:
|
|
56 |
return self.extract_reps(df).mul(self.policy_count, axis=0)
|
57 |
|
58 |
def compare(self, df, agg=None):
|
|
|
59 |
source = self.agg_by_cluster(df, agg)
|
60 |
target = self.extract_and_scale_reps(df, agg)
|
61 |
return pd.DataFrame({'actual': source.stack(), 'estimate':target.stack()})
|
62 |
|
63 |
def compare_total(self, df, agg=None):
|
|
|
64 |
if agg:
|
|
|
65 |
actual_values = {}
|
66 |
for col in df.columns:
|
67 |
if agg.get(col, 'sum') == 'mean':
|
68 |
actual_values[col] = df[col].mean()
|
69 |
-
else:
|
70 |
actual_values[col] = df[col].sum()
|
71 |
actual = pd.Series(actual_values)
|
72 |
|
|
|
73 |
reps_unscaled = self.extract_reps(df)
|
74 |
estimate_values = {}
|
75 |
|
76 |
for col in df.columns:
|
77 |
if agg.get(col, 'sum') == 'mean':
|
|
|
78 |
weighted_sum = (reps_unscaled[col] * self.policy_count).sum()
|
79 |
total_weight = self.policy_count.sum()
|
80 |
estimate_values[col] = weighted_sum / total_weight if total_weight > 0 else 0
|
81 |
-
else:
|
82 |
estimate_values[col] = (reps_unscaled[col] * self.policy_count).sum()
|
83 |
|
84 |
estimate = pd.Series(estimate_values)
|
85 |
-
|
|
|
86 |
actual = df.sum()
|
87 |
estimate = self.extract_and_scale_reps(df).sum()
|
88 |
|
|
|
89 |
error = np.where(actual != 0, estimate / actual - 1, 0)
|
|
|
90 |
return pd.DataFrame({'actual': actual, 'estimate': estimate, 'error': error})
|
91 |
|
|
|
92 |
def plot_cashflows_comparison(cfs_list, cluster_obj, titles):
|
|
|
93 |
if not cfs_list or not cluster_obj or not titles:
|
94 |
return None
|
95 |
-
|
96 |
num_plots = len(cfs_list)
|
97 |
if num_plots == 0:
|
98 |
return None
|
99 |
|
|
|
100 |
cols = 2
|
101 |
rows = (num_plots + cols - 1) // cols
|
102 |
|
@@ -106,17 +120,11 @@ def plot_cashflows_comparison(cfs_list, cluster_obj, titles):
|
|
106 |
for i, (df, title) in enumerate(zip(cfs_list, titles)):
|
107 |
if i < len(axes):
|
108 |
comparison = cluster_obj.compare_total(df)
|
109 |
-
|
110 |
-
# Plot using seaborn lineplot for cleaner aesthetics
|
111 |
-
data_to_plot = comparison[['actual', 'estimate']].reset_index()
|
112 |
-
data_melted = data_to_plot.melt(id_vars='index', var_name='Type', value_name='Value')
|
113 |
-
|
114 |
-
sns.lineplot(data=data_melted, x='index', y='Value', hue='Type', ax=axes[i])
|
115 |
-
axes[i].set_title(title)
|
116 |
axes[i].set_xlabel('Time')
|
117 |
axes[i].set_ylabel('Value')
|
118 |
-
axes[i].grid(True)
|
119 |
|
|
|
120 |
for j in range(i + 1, len(axes)):
|
121 |
fig.delaxes(axes[j])
|
122 |
|
@@ -129,7 +137,9 @@ def plot_cashflows_comparison(cfs_list, cluster_obj, titles):
|
|
129 |
return img
|
130 |
|
131 |
def plot_scatter_comparison(df_compare_output, title):
|
|
|
132 |
if df_compare_output is None or df_compare_output.empty:
|
|
|
133 |
fig, ax = plt.subplots(figsize=(12, 8))
|
134 |
ax.text(0.5, 0.5, "No data to display", ha='center', va='center', fontsize=15)
|
135 |
ax.set_title(title)
|
@@ -143,20 +153,17 @@ def plot_scatter_comparison(df_compare_output, title):
|
|
143 |
fig, ax = plt.subplots(figsize=(12, 8))
|
144 |
|
145 |
if not isinstance(df_compare_output.index, pd.MultiIndex) or df_compare_output.index.nlevels < 2:
|
146 |
-
|
147 |
-
|
148 |
else:
|
149 |
-
# Prepare data for seaborn
|
150 |
-
plot_data = df_compare_output.reset_index()
|
151 |
-
level_1_name = df_compare_output.index.names[1]
|
152 |
-
|
153 |
unique_levels = df_compare_output.index.get_level_values(1).unique()
|
|
|
154 |
|
|
|
|
|
|
|
155 |
if len(unique_levels) > 1 and len(unique_levels) <= 10:
|
156 |
-
|
157 |
-
data=plot_data, s=9, alpha=0.6, ax=ax)
|
158 |
-
else:
|
159 |
-
sns.scatterplot(x='actual', y='estimate', data=plot_data, s=9, alpha=0.6, ax=ax)
|
160 |
|
161 |
ax.set_xlabel('Actual')
|
162 |
ax.set_ylabel('Estimate')
|
@@ -169,9 +176,9 @@ def plot_scatter_comparison(df_compare_output, title):
|
|
169 |
np.max([ax.get_xlim(), ax.get_ylim()]),
|
170 |
]
|
171 |
if lims[0] != lims[1]:
|
172 |
-
|
173 |
-
|
174 |
-
|
175 |
|
176 |
buf = io.BytesIO()
|
177 |
plt.savefig(buf, format='png', dpi=100)
|
@@ -180,15 +187,18 @@ def plot_scatter_comparison(df_compare_output, title):
|
|
180 |
plt.close(fig)
|
181 |
return img
|
182 |
|
|
|
183 |
def process_files(cashflow_base_path, cashflow_lapse_path, cashflow_mort_path,
|
184 |
policy_data_path, pv_base_path, pv_lapse_path, pv_mort_path):
|
|
|
185 |
try:
|
186 |
-
# Read files
|
187 |
cfs = pd.read_excel(cashflow_base_path, index_col=0)
|
188 |
cfs_lapse50 = pd.read_excel(cashflow_lapse_path, index_col=0)
|
189 |
cfs_mort15 = pd.read_excel(cashflow_mort_path, index_col=0)
|
190 |
|
191 |
pol_data_full = pd.read_excel(policy_data_path, index_col=0)
|
|
|
192 |
required_cols = ['age_at_entry', 'policy_term', 'sum_assured', 'duration_mth']
|
193 |
if all(col in pol_data_full.columns for col in required_cols):
|
194 |
pol_data = pol_data_full[required_cols]
|
@@ -204,90 +214,104 @@ def process_files(cashflow_base_path, cashflow_lapse_path, cashflow_mort_path,
|
|
204 |
scen_titles = ['Base', 'Lapse+50%', 'Mort+15%']
|
205 |
|
206 |
results = {}
|
|
|
207 |
mean_attrs = {'age_at_entry':'mean', 'policy_term':'mean', 'duration_mth':'mean', 'sum_assured': 'sum'}
|
208 |
|
209 |
-
# Cashflow Calibration
|
210 |
cluster_cfs = Clusters(cfs)
|
211 |
-
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
|
220 |
-
|
221 |
-
|
|
|
|
|
222 |
if not pol_data.empty and (pol_data.max() - pol_data.min()).all() != 0:
|
223 |
-
|
224 |
else:
|
225 |
gr.Warning("Policy data for attribute calibration is empty or has no variance. Skipping attribute calibration plots.")
|
226 |
loc_vars_attrs = pol_data
|
227 |
|
228 |
if not loc_vars_attrs.empty:
|
229 |
cluster_attrs = Clusters(loc_vars_attrs)
|
230 |
-
results.
|
231 |
-
|
232 |
-
|
233 |
-
|
234 |
-
|
235 |
-
'attr_scatter_cashflows_base': plot_scatter_comparison(cluster_attrs.compare(cfs), 'Policy Attr. Calib. - Cashflows (Base)')
|
236 |
-
})
|
237 |
else:
|
238 |
-
results.
|
239 |
-
|
240 |
-
|
241 |
-
|
242 |
-
|
243 |
-
|
244 |
-
|
245 |
-
|
246 |
-
# Present Value Calibration
|
247 |
cluster_pvs = Clusters(pvs)
|
248 |
-
|
249 |
-
|
250 |
-
|
251 |
-
|
252 |
-
|
253 |
-
|
254 |
-
|
255 |
-
|
256 |
-
|
257 |
-
|
258 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
259 |
def get_error_safe(compare_result, col_name=None):
|
260 |
if compare_result.empty:
|
261 |
return np.nan
|
262 |
if col_name and col_name in compare_result.index:
|
263 |
return abs(compare_result.loc[col_name, 'error'])
|
264 |
else:
|
|
|
265 |
return abs(compare_result['error']).mean()
|
266 |
|
|
|
267 |
key_pv_col = None
|
268 |
for potential_col in ['PV_NetCF', 'pv_net_cf', 'net_cf_pv', 'PV_Net_CF']:
|
269 |
if potential_col in pvs.columns:
|
270 |
key_pv_col = potential_col
|
271 |
break
|
272 |
|
273 |
-
|
274 |
-
|
275 |
-
|
276 |
-
|
277 |
-
|
278 |
-
|
279 |
-
|
280 |
-
|
281 |
-
|
282 |
-
|
283 |
-
|
284 |
-
|
285 |
-
get_error_safe(
|
286 |
-
get_error_safe(cluster_pvs.compare_total(pvs_lapse50), key_pv_col),
|
287 |
-
get_error_safe(cluster_pvs.compare_total(pvs_mort15), key_pv_col)
|
288 |
]
|
289 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
290 |
|
|
|
291 |
summary_df = pd.DataFrame(error_data, index=['Base', 'Lapse+50%', 'Mort+15%'])
|
292 |
|
293 |
fig_summary, ax_summary = plt.subplots(figsize=(10, 6))
|
@@ -317,6 +341,7 @@ def process_files(cashflow_base_path, cashflow_lapse_path, cashflow_mort_path,
|
|
317 |
gr.Error(f"Error processing files: {str(e)}")
|
318 |
return {"error": f"Error processing files: {str(e)}"}
|
319 |
|
|
|
320 |
def create_interface():
|
321 |
with gr.Blocks(title="Cluster Model Points Analysis") as demo:
|
322 |
gr.Markdown("""
|
@@ -394,30 +419,37 @@ def create_interface():
|
|
394 |
pv_total_pv_lapse_out = gr.Dataframe(label="PVs - Lapse Stress Total")
|
395 |
pv_total_pv_mort_out = gr.Dataframe(label="PVs - Mortality Stress Total")
|
396 |
|
|
|
397 |
def get_all_output_components():
|
398 |
return [
|
399 |
summary_plot_output,
|
|
|
400 |
cf_total_base_table_out, cf_policy_attrs_total_out,
|
401 |
cf_cashflow_plot_out, cf_scatter_cashflows_base_out,
|
402 |
cf_pv_total_base_out, cf_pv_total_lapse_out, cf_pv_total_mort_out,
|
|
|
403 |
attr_total_cf_base_out, attr_policy_attrs_total_out,
|
404 |
attr_cashflow_plot_out, attr_scatter_cashflows_base_out, attr_total_pv_base_out,
|
|
|
405 |
pv_total_cf_base_out, pv_policy_attrs_total_out,
|
406 |
pv_cashflow_plot_out, pv_scatter_pvs_base_out,
|
407 |
pv_total_pv_base_out, pv_total_pv_lapse_out, pv_total_pv_mort_out
|
408 |
]
|
409 |
|
|
|
410 |
def handle_analysis(f1, f2, f3, f4, f5, f6, f7):
|
411 |
files = [f1, f2, f3, f4, f5, f6, f7]
|
412 |
-
file_paths = []
|
413 |
|
|
|
414 |
for i, f_obj in enumerate(files):
|
415 |
if f_obj is None:
|
416 |
gr.Error(f"Missing file input for argument {i+1}. Please upload all files or load examples.")
|
417 |
return [None] * len(get_all_output_components())
|
418 |
|
|
|
419 |
if hasattr(f_obj, 'name') and isinstance(f_obj.name, str):
|
420 |
file_paths.append(f_obj.name)
|
|
|
421 |
elif isinstance(f_obj, str):
|
422 |
file_paths.append(f_obj)
|
423 |
else:
|
@@ -431,11 +463,14 @@ def create_interface():
|
|
431 |
|
432 |
return [
|
433 |
results.get('summary_plot'),
|
|
|
434 |
results.get('cf_total_base_table'), results.get('cf_policy_attrs_total'),
|
435 |
results.get('cf_cashflow_plot'), results.get('cf_scatter_cashflows_base'),
|
436 |
results.get('cf_pv_total_base'), results.get('cf_pv_total_lapse'), results.get('cf_pv_total_mort'),
|
|
|
437 |
results.get('attr_total_cf_base'), results.get('attr_policy_attrs_total'),
|
438 |
results.get('attr_cashflow_plot'), results.get('attr_scatter_cashflows_base'), results.get('attr_total_pv_base'),
|
|
|
439 |
results.get('pv_total_cf_base'), results.get('pv_policy_attrs_total'),
|
440 |
results.get('pv_cashflow_plot'), results.get('pv_scatter_pvs_base'),
|
441 |
results.get('pv_total_pv_base'), results.get('pv_total_pv_lapse'), results.get('pv_total_pv_mort')
|
@@ -448,6 +483,7 @@ def create_interface():
|
|
448 |
outputs=get_all_output_components()
|
449 |
)
|
450 |
|
|
|
451 |
def load_example_files():
|
452 |
missing_files = [fp for fp in EXAMPLE_FILES.values() if not os.path.exists(fp)]
|
453 |
if missing_files:
|
|
|
2 |
import numpy as np
|
3 |
import pandas as pd
|
4 |
from sklearn.cluster import KMeans
|
5 |
+
from sklearn.metrics import pairwise_distances_argmin_min, r2_score
|
|
|
6 |
import matplotlib.pyplot as plt
|
7 |
+
import matplotlib.cm
|
8 |
import io
|
9 |
+
import os # Added for path joining
|
10 |
from PIL import Image
|
11 |
|
12 |
# Define the paths for example data
|
|
|
23 |
|
24 |
class Clusters:
|
25 |
def __init__(self, loc_vars):
|
26 |
+
self.kmeans = kmeans = KMeans(n_clusters=1000, random_state=0, n_init=10).fit(np.ascontiguousarray(loc_vars))
|
27 |
+
closest, _ = pairwise_distances_argmin_min(kmeans.cluster_centers_, np.ascontiguousarray(loc_vars))
|
28 |
|
29 |
+
rep_ids = pd.Series(data=(closest+1)) # 0-based to 1-based indexes
|
30 |
rep_ids.name = 'policy_id'
|
31 |
rep_ids.index.name = 'cluster_id'
|
32 |
self.rep_ids = rep_ids
|
|
|
34 |
self.policy_count = self.agg_by_cluster(pd.DataFrame({'policy_count': [1] * len(loc_vars)}))['policy_count']
|
35 |
|
36 |
def agg_by_cluster(self, df, agg=None):
|
37 |
+
"""Aggregate columns by cluster"""
|
38 |
temp = df.copy()
|
39 |
temp['cluster_id'] = self.kmeans.labels_
|
40 |
temp = temp.set_index('cluster_id')
|
|
|
42 |
return temp.groupby(temp.index).agg(agg)
|
43 |
|
44 |
def extract_reps(self, df):
|
45 |
+
"""Extract the rows of representative policies"""
|
46 |
temp = pd.merge(self.rep_ids, df.reset_index(), how='left', on='policy_id')
|
47 |
temp.index.name = 'cluster_id'
|
48 |
return temp.drop('policy_id', axis=1)
|
49 |
|
50 |
def extract_and_scale_reps(self, df, agg=None):
|
51 |
+
"""Extract and scale the rows of representative policies"""
|
52 |
if agg:
|
53 |
cols = df.columns
|
54 |
mult = pd.DataFrame({c: (self.policy_count if (c not in agg or agg[c] == 'sum') else 1) for c in cols})
|
55 |
+
# Ensure mult has same index as extract_reps(df) for proper alignment
|
56 |
extracted_df = self.extract_reps(df)
|
57 |
mult.index = extracted_df.index
|
58 |
return extracted_df.mul(mult)
|
|
|
60 |
return self.extract_reps(df).mul(self.policy_count, axis=0)
|
61 |
|
62 |
def compare(self, df, agg=None):
|
63 |
+
"""Returns a multi-indexed Dataframe comparing actual and estimate"""
|
64 |
source = self.agg_by_cluster(df, agg)
|
65 |
target = self.extract_and_scale_reps(df, agg)
|
66 |
return pd.DataFrame({'actual': source.stack(), 'estimate':target.stack()})
|
67 |
|
68 |
def compare_total(self, df, agg=None):
|
69 |
+
"""Aggregate df by columns"""
|
70 |
if agg:
|
71 |
+
# Calculate actual values using specified aggregation
|
72 |
actual_values = {}
|
73 |
for col in df.columns:
|
74 |
if agg.get(col, 'sum') == 'mean':
|
75 |
actual_values[col] = df[col].mean()
|
76 |
+
else: # sum
|
77 |
actual_values[col] = df[col].sum()
|
78 |
actual = pd.Series(actual_values)
|
79 |
|
80 |
+
# Calculate estimate values
|
81 |
reps_unscaled = self.extract_reps(df)
|
82 |
estimate_values = {}
|
83 |
|
84 |
for col in df.columns:
|
85 |
if agg.get(col, 'sum') == 'mean':
|
86 |
+
# Weighted average for mean columns
|
87 |
weighted_sum = (reps_unscaled[col] * self.policy_count).sum()
|
88 |
total_weight = self.policy_count.sum()
|
89 |
estimate_values[col] = weighted_sum / total_weight if total_weight > 0 else 0
|
90 |
+
else: # sum
|
91 |
estimate_values[col] = (reps_unscaled[col] * self.policy_count).sum()
|
92 |
|
93 |
estimate = pd.Series(estimate_values)
|
94 |
+
|
95 |
+
else: # Original logic if no agg is specified (all sum)
|
96 |
actual = df.sum()
|
97 |
estimate = self.extract_and_scale_reps(df).sum()
|
98 |
|
99 |
+
# Calculate error, handling division by zero
|
100 |
error = np.where(actual != 0, estimate / actual - 1, 0)
|
101 |
+
|
102 |
return pd.DataFrame({'actual': actual, 'estimate': estimate, 'error': error})
|
103 |
|
104 |
+
|
105 |
def plot_cashflows_comparison(cfs_list, cluster_obj, titles):
|
106 |
+
"""Create cashflow comparison plots"""
|
107 |
if not cfs_list or not cluster_obj or not titles:
|
108 |
return None
|
|
|
109 |
num_plots = len(cfs_list)
|
110 |
if num_plots == 0:
|
111 |
return None
|
112 |
|
113 |
+
# Determine subplot layout
|
114 |
cols = 2
|
115 |
rows = (num_plots + cols - 1) // cols
|
116 |
|
|
|
120 |
for i, (df, title) in enumerate(zip(cfs_list, titles)):
|
121 |
if i < len(axes):
|
122 |
comparison = cluster_obj.compare_total(df)
|
123 |
+
comparison[['actual', 'estimate']].plot(ax=axes[i], grid=True, title=title)
|
|
|
|
|
|
|
|
|
|
|
|
|
124 |
axes[i].set_xlabel('Time')
|
125 |
axes[i].set_ylabel('Value')
|
|
|
126 |
|
127 |
+
# Hide any unused subplots
|
128 |
for j in range(i + 1, len(axes)):
|
129 |
fig.delaxes(axes[j])
|
130 |
|
|
|
137 |
return img
|
138 |
|
139 |
def plot_scatter_comparison(df_compare_output, title):
|
140 |
+
"""Create scatter plot comparison from compare() output"""
|
141 |
if df_compare_output is None or df_compare_output.empty:
|
142 |
+
# Create a blank plot with a message
|
143 |
fig, ax = plt.subplots(figsize=(12, 8))
|
144 |
ax.text(0.5, 0.5, "No data to display", ha='center', va='center', fontsize=15)
|
145 |
ax.set_title(title)
|
|
|
153 |
fig, ax = plt.subplots(figsize=(12, 8))
|
154 |
|
155 |
if not isinstance(df_compare_output.index, pd.MultiIndex) or df_compare_output.index.nlevels < 2:
|
156 |
+
gr.Warning("Scatter plot data is not in the expected multi-index format. Plotting raw actual vs estimate.")
|
157 |
+
ax.scatter(df_compare_output['actual'], df_compare_output['estimate'], s=9, alpha=0.6)
|
158 |
else:
|
|
|
|
|
|
|
|
|
159 |
unique_levels = df_compare_output.index.get_level_values(1).unique()
|
160 |
+
colors = matplotlib.cm.rainbow(np.linspace(0, 1, len(unique_levels)))
|
161 |
|
162 |
+
for item_level, color_val in zip(unique_levels, colors):
|
163 |
+
subset = df_compare_output.xs(item_level, level=1)
|
164 |
+
ax.scatter(subset['actual'], subset['estimate'], color=color_val, s=9, alpha=0.6, label=item_level)
|
165 |
if len(unique_levels) > 1 and len(unique_levels) <= 10:
|
166 |
+
ax.legend(title=df_compare_output.index.names[1])
|
|
|
|
|
|
|
167 |
|
168 |
ax.set_xlabel('Actual')
|
169 |
ax.set_ylabel('Estimate')
|
|
|
176 |
np.max([ax.get_xlim(), ax.get_ylim()]),
|
177 |
]
|
178 |
if lims[0] != lims[1]:
|
179 |
+
ax.plot(lims, lims, 'r-', linewidth=0.5)
|
180 |
+
ax.set_xlim(lims)
|
181 |
+
ax.set_ylim(lims)
|
182 |
|
183 |
buf = io.BytesIO()
|
184 |
plt.savefig(buf, format='png', dpi=100)
|
|
|
187 |
plt.close(fig)
|
188 |
return img
|
189 |
|
190 |
+
|
191 |
def process_files(cashflow_base_path, cashflow_lapse_path, cashflow_mort_path,
|
192 |
policy_data_path, pv_base_path, pv_lapse_path, pv_mort_path):
|
193 |
+
"""Main processing function - now accepts file paths"""
|
194 |
try:
|
195 |
+
# Read uploaded files using paths
|
196 |
cfs = pd.read_excel(cashflow_base_path, index_col=0)
|
197 |
cfs_lapse50 = pd.read_excel(cashflow_lapse_path, index_col=0)
|
198 |
cfs_mort15 = pd.read_excel(cashflow_mort_path, index_col=0)
|
199 |
|
200 |
pol_data_full = pd.read_excel(policy_data_path, index_col=0)
|
201 |
+
# Ensure the correct columns are selected for pol_data
|
202 |
required_cols = ['age_at_entry', 'policy_term', 'sum_assured', 'duration_mth']
|
203 |
if all(col in pol_data_full.columns for col in required_cols):
|
204 |
pol_data = pol_data_full[required_cols]
|
|
|
214 |
scen_titles = ['Base', 'Lapse+50%', 'Mort+15%']
|
215 |
|
216 |
results = {}
|
217 |
+
|
218 |
mean_attrs = {'age_at_entry':'mean', 'policy_term':'mean', 'duration_mth':'mean', 'sum_assured': 'sum'}
|
219 |
|
220 |
+
# --- 1. Cashflow Calibration ---
|
221 |
cluster_cfs = Clusters(cfs)
|
222 |
+
|
223 |
+
results['cf_total_base_table'] = cluster_cfs.compare_total(cfs)
|
224 |
+
results['cf_policy_attrs_total'] = cluster_cfs.compare_total(pol_data, agg=mean_attrs)
|
225 |
+
|
226 |
+
results['cf_pv_total_base'] = cluster_cfs.compare_total(pvs)
|
227 |
+
results['cf_pv_total_lapse'] = cluster_cfs.compare_total(pvs_lapse50)
|
228 |
+
results['cf_pv_total_mort'] = cluster_cfs.compare_total(pvs_mort15)
|
229 |
+
|
230 |
+
results['cf_cashflow_plot'] = plot_cashflows_comparison(cfs_list, cluster_cfs, scen_titles)
|
231 |
+
results['cf_scatter_cashflows_base'] = plot_scatter_comparison(cluster_cfs.compare(cfs), 'Cashflow Calib. - Cashflows (Base)')
|
232 |
+
|
233 |
+
# --- 2. Policy Attribute Calibration ---
|
234 |
+
# Standardize policy attributes
|
235 |
if not pol_data.empty and (pol_data.max() - pol_data.min()).all() != 0:
|
236 |
+
loc_vars_attrs = (pol_data - pol_data.min()) / (pol_data.max() - pol_data.min())
|
237 |
else:
|
238 |
gr.Warning("Policy data for attribute calibration is empty or has no variance. Skipping attribute calibration plots.")
|
239 |
loc_vars_attrs = pol_data
|
240 |
|
241 |
if not loc_vars_attrs.empty:
|
242 |
cluster_attrs = Clusters(loc_vars_attrs)
|
243 |
+
results['attr_total_cf_base'] = cluster_attrs.compare_total(cfs)
|
244 |
+
results['attr_policy_attrs_total'] = cluster_attrs.compare_total(pol_data, agg=mean_attrs)
|
245 |
+
results['attr_total_pv_base'] = cluster_attrs.compare_total(pvs)
|
246 |
+
results['attr_cashflow_plot'] = plot_cashflows_comparison(cfs_list, cluster_attrs, scen_titles)
|
247 |
+
results['attr_scatter_cashflows_base'] = plot_scatter_comparison(cluster_attrs.compare(cfs), 'Policy Attr. Calib. - Cashflows (Base)')
|
|
|
|
|
248 |
else:
|
249 |
+
results['attr_total_cf_base'] = pd.DataFrame()
|
250 |
+
results['attr_policy_attrs_total'] = pd.DataFrame()
|
251 |
+
results['attr_total_pv_base'] = pd.DataFrame()
|
252 |
+
results['attr_cashflow_plot'] = None
|
253 |
+
results['attr_scatter_cashflows_base'] = None
|
254 |
+
|
255 |
+
# --- 3. Present Value Calibration ---
|
|
|
|
|
256 |
cluster_pvs = Clusters(pvs)
|
257 |
+
|
258 |
+
results['pv_total_cf_base'] = cluster_pvs.compare_total(cfs)
|
259 |
+
results['pv_policy_attrs_total'] = cluster_pvs.compare_total(pol_data, agg=mean_attrs)
|
260 |
+
|
261 |
+
results['pv_total_pv_base'] = cluster_pvs.compare_total(pvs)
|
262 |
+
results['pv_total_pv_lapse'] = cluster_pvs.compare_total(pvs_lapse50)
|
263 |
+
results['pv_total_pv_mort'] = cluster_pvs.compare_total(pvs_mort15)
|
264 |
+
|
265 |
+
results['pv_cashflow_plot'] = plot_cashflows_comparison(cfs_list, cluster_pvs, scen_titles)
|
266 |
+
results['pv_scatter_pvs_base'] = plot_scatter_comparison(cluster_pvs.compare(pvs), 'PV Calib. - PVs (Base)')
|
267 |
+
|
268 |
+
# --- Summary Comparison Plot Data ---
|
269 |
+
# Error metric for key PV column or mean absolute error
|
270 |
+
|
271 |
+
error_data = {}
|
272 |
+
|
273 |
+
# Function to safely get error value
|
274 |
def get_error_safe(compare_result, col_name=None):
|
275 |
if compare_result.empty:
|
276 |
return np.nan
|
277 |
if col_name and col_name in compare_result.index:
|
278 |
return abs(compare_result.loc[col_name, 'error'])
|
279 |
else:
|
280 |
+
# Use mean absolute error if specific column not found
|
281 |
return abs(compare_result['error']).mean()
|
282 |
|
283 |
+
# Determine key PV column (try common names)
|
284 |
key_pv_col = None
|
285 |
for potential_col in ['PV_NetCF', 'pv_net_cf', 'net_cf_pv', 'PV_Net_CF']:
|
286 |
if potential_col in pvs.columns:
|
287 |
key_pv_col = potential_col
|
288 |
break
|
289 |
|
290 |
+
# Cashflow Calibration Errors
|
291 |
+
error_data['CF Calib.'] = [
|
292 |
+
get_error_safe(cluster_cfs.compare_total(pvs), key_pv_col),
|
293 |
+
get_error_safe(cluster_cfs.compare_total(pvs_lapse50), key_pv_col),
|
294 |
+
get_error_safe(cluster_cfs.compare_total(pvs_mort15), key_pv_col)
|
295 |
+
]
|
296 |
+
|
297 |
+
# Policy Attribute Calibration Errors
|
298 |
+
if not loc_vars_attrs.empty:
|
299 |
+
error_data['Attr Calib.'] = [
|
300 |
+
get_error_safe(cluster_attrs.compare_total(pvs), key_pv_col),
|
301 |
+
get_error_safe(cluster_attrs.compare_total(pvs_lapse50), key_pv_col),
|
302 |
+
get_error_safe(cluster_attrs.compare_total(pvs_mort15), key_pv_col)
|
|
|
|
|
303 |
]
|
304 |
+
else:
|
305 |
+
error_data['Attr Calib.'] = [np.nan, np.nan, np.nan]
|
306 |
+
|
307 |
+
# Present Value Calibration Errors
|
308 |
+
error_data['PV Calib.'] = [
|
309 |
+
get_error_safe(cluster_pvs.compare_total(pvs), key_pv_col),
|
310 |
+
get_error_safe(cluster_pvs.compare_total(pvs_lapse50), key_pv_col),
|
311 |
+
get_error_safe(cluster_pvs.compare_total(pvs_mort15), key_pv_col)
|
312 |
+
]
|
313 |
|
314 |
+
# Create Summary Plot
|
315 |
summary_df = pd.DataFrame(error_data, index=['Base', 'Lapse+50%', 'Mort+15%'])
|
316 |
|
317 |
fig_summary, ax_summary = plt.subplots(figsize=(10, 6))
|
|
|
341 |
gr.Error(f"Error processing files: {str(e)}")
|
342 |
return {"error": f"Error processing files: {str(e)}"}
|
343 |
|
344 |
+
|
345 |
def create_interface():
|
346 |
with gr.Blocks(title="Cluster Model Points Analysis") as demo:
|
347 |
gr.Markdown("""
|
|
|
419 |
pv_total_pv_lapse_out = gr.Dataframe(label="PVs - Lapse Stress Total")
|
420 |
pv_total_pv_mort_out = gr.Dataframe(label="PVs - Mortality Stress Total")
|
421 |
|
422 |
+
# --- Helper function to prepare outputs ---
|
423 |
def get_all_output_components():
|
424 |
return [
|
425 |
summary_plot_output,
|
426 |
+
# Cashflow Calib Outputs
|
427 |
cf_total_base_table_out, cf_policy_attrs_total_out,
|
428 |
cf_cashflow_plot_out, cf_scatter_cashflows_base_out,
|
429 |
cf_pv_total_base_out, cf_pv_total_lapse_out, cf_pv_total_mort_out,
|
430 |
+
# Attribute Calib Outputs
|
431 |
attr_total_cf_base_out, attr_policy_attrs_total_out,
|
432 |
attr_cashflow_plot_out, attr_scatter_cashflows_base_out, attr_total_pv_base_out,
|
433 |
+
# PV Calib Outputs
|
434 |
pv_total_cf_base_out, pv_policy_attrs_total_out,
|
435 |
pv_cashflow_plot_out, pv_scatter_pvs_base_out,
|
436 |
pv_total_pv_base_out, pv_total_pv_lapse_out, pv_total_pv_mort_out
|
437 |
]
|
438 |
|
439 |
+
# --- Action for Analyze Button ---
|
440 |
def handle_analysis(f1, f2, f3, f4, f5, f6, f7):
|
441 |
files = [f1, f2, f3, f4, f5, f6, f7]
|
|
|
442 |
|
443 |
+
file_paths = []
|
444 |
for i, f_obj in enumerate(files):
|
445 |
if f_obj is None:
|
446 |
gr.Error(f"Missing file input for argument {i+1}. Please upload all files or load examples.")
|
447 |
return [None] * len(get_all_output_components())
|
448 |
|
449 |
+
# If f_obj is a Gradio FileData object (from direct upload)
|
450 |
if hasattr(f_obj, 'name') and isinstance(f_obj.name, str):
|
451 |
file_paths.append(f_obj.name)
|
452 |
+
# If f_obj is already a string path (from example load)
|
453 |
elif isinstance(f_obj, str):
|
454 |
file_paths.append(f_obj)
|
455 |
else:
|
|
|
463 |
|
464 |
return [
|
465 |
results.get('summary_plot'),
|
466 |
+
# CF Calib
|
467 |
results.get('cf_total_base_table'), results.get('cf_policy_attrs_total'),
|
468 |
results.get('cf_cashflow_plot'), results.get('cf_scatter_cashflows_base'),
|
469 |
results.get('cf_pv_total_base'), results.get('cf_pv_total_lapse'), results.get('cf_pv_total_mort'),
|
470 |
+
# Attr Calib
|
471 |
results.get('attr_total_cf_base'), results.get('attr_policy_attrs_total'),
|
472 |
results.get('attr_cashflow_plot'), results.get('attr_scatter_cashflows_base'), results.get('attr_total_pv_base'),
|
473 |
+
# PV Calib
|
474 |
results.get('pv_total_cf_base'), results.get('pv_policy_attrs_total'),
|
475 |
results.get('pv_cashflow_plot'), results.get('pv_scatter_pvs_base'),
|
476 |
results.get('pv_total_pv_base'), results.get('pv_total_pv_lapse'), results.get('pv_total_pv_mort')
|
|
|
483 |
outputs=get_all_output_components()
|
484 |
)
|
485 |
|
486 |
+
# --- Action for Load Example Data Button ---
|
487 |
def load_example_files():
|
488 |
missing_files = [fp for fp in EXAMPLE_FILES.values() if not os.path.exists(fp)]
|
489 |
if missing_files:
|