You don't need 4 hours a day. You need one focused hour, done the same way every day.
Same dataset all week. Only the concept changes. Sunday, you ship a small project.
Start today, not Monday.
How the hour splits
| Minutes | Do this |
|---|---|
| 0-10 | Retype yesterday's code from memory |
| 10-30 | Learn one new pandas idea |
| 30-50 | Use it on a real CSV |
| 50-60 | 3 lines of notes, push to GitHub |
Same dataset all week. Only the concept changes.
Mon to Sat: one idea a day
- Mon
Read and filter
Load the file, keep the rows you need.
df = pd.read_csv('sales.csv') df[df['city'] == 'Indore'] - Tue
groupby
Totals per city. Asked in almost every test.
df.groupby('city')['amount'].sum() - Wed
merge
Like a SQL LEFT JOIN.
orders.merge(customers, on='cust_id', how='left') - Thu
Dates
Convert first, then pull the month.
df['order_date'] = pd.to_datetime(df['order_date']) df['month'] = df['order_date'].dt.to_period('M') - Fri
Cleaning
Duplicates and blanks, every single time.
df = df.drop_duplicates() df['amount'] = df['amount'].fillna(0) - Sat
pivot_table
City by month, like an Excel pivot.
pd.pivot_table(df, index='city', columns='month', values='amount', aggfunc='sum')
Sunday: one mini project
| 3 business questions | Top city, best month, repeat customers. Use only this week's code. |
| One dataset | A Kaggle sales CSV or your own monthly expense sheet. |
| Ship it | Push to GitHub with a README. 4 weeks = 4 small projects. |
Wasted hour vs useful hour
Wasted
- 3-hour tutorial on 2x speed
- Copying code without running it
- New topic daily, no revision
Useful
- Type every line yourself
- Same dataset for a full week
- First 10 minutes = revision
Want the 10 pandas skills this builds toward? See Python skills for data jobs.