200/main.py
2025-11-27 04:48:20 +00:00

152 lines
4.4 KiB
Python

#!/usr/bin/env python
# coding: utf-8
# In[3]:
import pandas as pd
# In[4]:
from tms_data_interface import SQLQueryInterface
seq = SQLQueryInterface(schema="transactionschema")
# In[5]:
seq.execute_raw("show tables")
# In[6]:
query = """
select final.CUSTOMER_NUMBER_main as Focal_id,
final.Credit_transaction_amount,
final.SEGMENT,
final.RISK,
final.SAR_FLAG
from
(
(
select subquery.CUSTOMER_NUMBER_1 as CUSTOMER_NUMBER_main,
subquery.Credit_transaction_amount
from
(
(
select customer_number as CUSTOMER_NUMBER_1,
sum(transaction_amount) as Credit_transaction_amount
from
(
select *
from {trans_data} as trans_table
left join {acc_data} as acc_table
on trans_table.benef_account_number = acc_table.account_number
where trans_table.transaction_desc = 'WIRE RELATED TRANSACTION'
)
where account_number not in ('None')
group by 1
) credit
) subquery
) main left join
(
select subquery.CUSTOMER_NUMBER_3 as CUSTOMER_NUMBER_cust,
subquery.SEGMENT,
subquery.RISK,
case
when subquery.SAR_FLAG is NULL then 'N'
else subquery.SAR_FLAG
end as SAR_FLAG
from
(
(
select customer_number as CUSTOMER_NUMBER_3,
business_segment as SEGMENT,
case
when RISK_CLASSIFICATION = 1 then 'Low Risk'
when RISK_CLASSIFICATION = 2 then 'Medium Risk'
when RISK_CLASSIFICATION = 3 then 'High Risk'
else 'Unknown Risk'
end AS RISK
from {cust_data}
) cd left join
(
select customer_number as CUSTOMER_NUMBER_4,
sar_flag as SAR_FLAG
from {alert_data}
) ad on cd.CUSTOMER_NUMBER_3 = ad.CUSTOMER_NUMBER_4
) subquery
) cust_alert on cust_alert.CUSTOMER_NUMBER_cust = main.CUSTOMER_NUMBER_main
) final
"""
# In[25]:
from tms_data_interface import SQLQueryInterface
class Scenario:
seq = SQLQueryInterface(schema="transactionschema")
def logic(self, **kwargs):
row_list = self.seq.execute_raw(query.format(trans_data="transaction10m",
cust_data="customer_data_v1",
acc_data="account_data_v1",
alert_data="alert_data_v1")
)
cols = ["Focal_id", "Total_Wire_Deposit_Amt",
"Segment", "Risk", "SAR_FLAG"]
df = pd.DataFrame(row_list, columns = cols)
df['Total_Wire_Deposit_Amt'] = df['Total_Wire_Deposit_Amt'].astype('int')
# df['Segment'] = 'Individual'
p98 = (
df.groupby("Segment")["Total_Wire_Deposit_Amt"]
.quantile(0.98)
.reset_index()
.rename(columns={"Total_Wire_Deposit_Amt": "P98_Value"})
)
print(p98)
# Merge percentile back to main dataframe
df = df.merge(p98, on="Segment", how="left")
# Step 2: Identify population above 98th percentile
high_pop = df["Total_Wire_Deposit_Amt"] > df["P98_Value"]
print(high_pop)
# Step 3: From this high-risk population, select 0.1% random sample
sample_fraction = 0.1 # 0.1%
high_pop_indices = df[high_pop].sample(frac=sample_fraction, random_state=42).index
# Step 4: Assign SAR_FLAG
df["SAR_FLAG"] = "N" # default for all
df.loc[high_pop_indices, "SAR_FLAG"] = "Y" # assign Y to random 0.1% above 98th percentile
return df
# In[28]:
# sen = Scenario()
# a = sen.logic()
# In[29]:
# a[a['SAR_FLAG'] == "Y"]
# In[ ]:
#tst cmt