Aggregation -- Weekly Average Temperature

The Analogy

You recorded the temperature every hour for a whole week. That's 168 readings! Your teacher asks: "What was the average temperature each day?"

Instead of reading all 168 numbers, you: 1. Group them by day (7 groups of 24 readings) 2. Summarise each group (add them up and divide by 24)

That's aggregation! You pick a window (one day) and a function (average) and let the computer crunch the numbers.

Available Functions

Function	What it does	Example
AVG	Average of all values	"Average temperature this hour"
MIN	Smallest value	"Coldest reading today"
MAX	Largest value	"Hottest reading today"
SUM	Total of all values	"Total rainfall this week"
COUNT	Number of readings	"How many readings per hour?"

In Code

from datetime import timedelta
from pytsdb import TimeSeriesDB, AggFunc

db = TimeSeriesDB()
db.create_series("temp")
# ... add hourly readings ...

# Get the average temperature per hour
hourly_avg = db.aggregate("temp", window=timedelta(hours=1), func=AggFunc.AVG)

# Get the max temperature per day
daily_max = db.aggregate("temp", window=timedelta(days=1), func=AggFunc.MAX)

How Time Buckets Work

Every data point's timestamp is mapped to a bucket. For a 1-hour window, all points between 10:00 and 10:59 go into the "10:00" bucket. The function is then applied to each bucket independently.