如何用Python识别日期结构并检查用户周活跃频次（含容错）

Hey there! Let's break down how to solve your two Python/pandas problems with clear, actionable code tailored to your DataFrame.

1. Identifying & Processing Date Structures #

First off, your just_dates column looks like string-formatted dates. To work with weeks and date-based logic, we need to convert this column to pandas' datetime type first—this unlocks all the date-specific tools we need.

Step 1: Convert to Datetime Type #

Use pd.to_datetime() to parse the strings into proper datetime objects. Pandas will automatically infer most standard date formats, but you can specify a format explicitly if you want better performance:

import pandas as pd

# Convert the date column to datetime (auto-infer format)
df['just_dates'] = pd.to_datetime(df['just_dates'], infer_datetime_format=True)

# Verify the conversion worked
print(df['just_dates'].dtype)  # Should output `datetime64[ns]`

If your dates follow a strict format (like YYYY-MM-DD), add format='%Y-%m-%d' to the pd.to_datetime() call to speed things up.

Step 2: Extract Week Information #

Once we have datetime objects, we can pull out year and week numbers (using ISO standard, where weeks start on Monday) to group data by week later:

# Add year and week columns (ISO standard)
df['year'] = df['just_dates'].dt.isocalendar().year
df['week'] = df['just_dates'].dt.isocalendar().week

2. Check Weekly Activity (At Least 2 Active Days) + Fault Tolerance #

Now let's build the logic to check if each week has at least 2 active days, plus a way to exempt specific time periods from this requirement.

Step 1: Calculate Weekly Active Days #

First, group the data by year and week to count how many active days each week has:

# Group by year + week, count active days per week
weekly_activity = df.groupby(['year', 'week'])['just_dates'].count().reset_index(name='active_days')

# Initial check: does the week meet the 2+ active days requirement?
weekly_activity['meets_requirement'] = weekly_activity['active_days'] >= 2

Step 2: Add Fault Tolerance for Specific Periods #

Let's implement two common ways to add exceptions: exempting specific weeks, or exempting a date range.

Option 1: Exempt Specific Weeks

Define a list of weeks (as (year, week) tuples) that don't need to meet the 2-day requirement, then adjust the meets_requirement flag:

# Define weeks to exempt (example: 2017 week 23, your "second-to-last week")
fault_tolerant_weeks = [(2017, 23)]

# Mark which weeks are exempt
weekly_activity['is_fault_tolerant'] = weekly_activity.apply(
    lambda row: (row['year'], row['week']) in fault_tolerant_weeks, axis=1
)

# Update the requirement check: exempt weeks automatically pass
weekly_activity['meets_requirement'] = weekly_activity.apply(
    lambda row: True if row['is_fault_tolerant'] else (row['active_days'] >= 2),
    axis=1
)

Option 2: Exempt a Date Range

If you need to exempt a specific date window (like a holiday period), calculate the start/end of each week and check if it overlaps with your exempt range:

# Calculate start and end dates for each week (ISO: Monday to Sunday)
weekly_activity['week_start'] = weekly_activity.apply(
    lambda row: pd.to_datetime(f"{row['year']}-W{row['week']}-1", format='%G-W%V-%u'),
    axis=1
)
weekly_activity['week_end'] = weekly_activity['week_start'] + pd.Timedelta(days=6)

# Define your exempt date range
fault_tolerant_start = pd.to_datetime('2017-06-01')
fault_tolerant_end = pd.to_datetime('2017-06-03')

# Check if the week overlaps with the exempt range
weekly_activity['is_fault_tolerant'] = weekly_activity.apply(
    lambda row: (row['week_start'] <= fault_tolerant_end) & (row['week_end'] >= fault_tolerant_start),
    axis=1
)

# Update the requirement check
weekly_activity['meets_requirement'] = weekly_activity.apply(
    lambda row: True if row['is_fault_tolerant'] else (row['active_days'] >= 2),
    axis=1
)

Step 3: Merge Results Back to Original DataFrame #

If you want each row in your original df to show whether its week meets the requirement (or is exempt), merge the weekly results back:

df = df.merge(
    weekly_activity[['year', 'week', 'meets_requirement', 'is_fault_tolerant']],
    on=['year', 'week'],
    how='left'
)

Verify Your Example #

For your sample data:

• The first week (2015 week 23) has 5 active days → meets_requirement=True
• The second-to-last week (2017 week 23) has 1 active day → if marked as exempt, meets_requirement=True; otherwise False
• The last week has 3 active days → meets_requirement=True

内容的提问来源于stack exchange，提问作者Dennis