-- A quick morning check on a key metric table
SELECT event_date, COUNT(*) AS rows_loaded,
       COUNT(DISTINCT user_id) AS users
FROM daily_user_events
WHERE event_date >= CURRENT_DATE - INTERVAL '7 days'
GROUP BY event_date
ORDER BY event_date;
-- A sudden drop may mean a pipeline issue, not a real change in users

Daily planner
Top priority: ______________
Questions to clarify with stakeholders: ______________
Data tasks: ______________
Analysis/modelling task: ______________
What I will share today: ______________
Blockers: ______________

import pandas as pd

orders = pd.read_csv("orders.csv", parse_dates=["order_date"])
customers = pd.read_csv("customers.csv")

df = orders.merge(customers, on="customer_id", how="left", validate="many_to_one")
df["city"] = df["city"].fillna("Unknown").str.title()
weekly = df.set_index("order_date").groupby("city").resample("W")["amount"].sum()
print(weekly.groupby(level=0).tail(2))

from sklearn.dummy import DummyClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import cross_val_score
from sklearn.datasets import load_breast_cancer

X, y = load_breast_cancer(return_X_y=True)
base = cross_val_score(DummyClassifier(strategy="most_frequent"), X, y, cv=5).mean()
model = cross_val_score(LogisticRegression(max_iter=5000), X, y, cv=5).mean()
print(f"baseline {base:.3f} vs model {model:.3f}")

Findings: Why did weekly orders dip?
Answer: The dip came almost entirely from one city after a delivery partner outage.
Evidence: weekly orders by city chart; other cities stable within normal range.
Confidence: high — timing matches the outage log.
Recommendation: add a backup partner for that city; monitor recovery next week.
Details: notebook link

A typical weekly rhythm
Mon: plan the week, align priorities with stakeholders
Tue-Thu: deep work on analysis or modelling
Thu: share interim results for feedback
Fri: write-up, code review, documentation, learning time

Project kick-off notes
Business question: ______________
Decision it will inform: ______________
Success metric: ______________
Data sources: ______________
Deadline / check-ins: ______________
Stakeholders: ______________
Risks (data quality, access): ______________

Analysis checklist
☐ Question written down and agreed
☐ Data checked for duplicates, nulls, date ranges
☐ Totals reconciled with a trusted source
☐ Charts labelled and honest (axes, units)
☐ Baseline or comparison included
☐ Conclusion and recommendation written
☐ Limitations noted

Weekly learning time (1-2 hours)
- Read one engineering or statistics article
- Try one new library function on real data
- Review a teammate's notebook
- Note one thing to apply next week