Which hour is the busiest?

The manager asks: "Which hour is the busiest? Who books the most?"

The club has 300 bookings in a file. Counting by hand takes hours. A short program does it in a second.

In the project you may be given a data file and asked questions like these.

Key words

Word	What it means	Example
CSV	A text file where each line is a record and commas separate the fields.	Sam,1,10
Dataset	A collection of data about one subject.	all the bookings
Data frame	A table of data inside pandas, with rows and columns.	one row per booking
Library	Code somebody else wrote that you can use.	pandas
Trend	A pattern that goes up or down over time.	more bookings each week
Average	The total divided by how many values there are.	(12 + 8 + 10) divided by 3 is 10
Axis	A line of a chart with a scale and a label.	hour along the bottom
Legend	A small box that says what each colour or line means.	blue is Court 1

The cells students fill in are left blank in this preview.

Try it: count with plain Python

Code: python

text = "member,court,hour\nSam,1,10\nSam,2,11\nAisha,1,10\nLee,2,10\n"

lines = text.strip().split("\n")
counts = {}
for i in range(1, len(lines)):    # start at 1 to skip the header
    hour = lines[i].split(",")[2]
    if hour in counts:
        counts[hour] = counts[hour] + 1
    else:
        counts[hour] = 1

print(counts)

# Hour 10 has 3 bookings.
# Add another line to the text and run it again.

Worked example: an average

Step	What you do	Result
1	Bookings on Monday, Tuesday and Wednesday	12, 8 and 10
2	Add them up	12 + 8 + 10 = 30
3	Count how many days	3
4	Divide the total by the count	30 divided by 3 = 10
5	Say it in words	The club has 10 bookings a day on average

The cells students fill in are left blank in this preview.

The same job with pandas

Code: python

import pandas as pd

# bookings.csv has the columns: member, court, hour, price, week
data = pd.read_csv("bookings.csv")

print(data["hour"].value_counts())    # count each hour
print(data.groupby("member").size())  # bookings per member
court_one = data[data["court"] == 1]  # keep only the rows for court 1
print(len(court_one))                 # bookings on court 1
print(data["price"].sum())            # the total money
print(data["price"].mean())           # the average price

# A trend over time: bookings in each week, in week order.
print(data.groupby("week").size())    # a rise or fall from week to week is a trend

Where pandas runs

data-callouts.note.label

The pandas and matplotlib blocks are for reading. The lesson window has only plain Python. Run them in the Python set-up your lecturer gives you. The tasks in this lesson use plain Python so they run anywhere.

A chart with matplotlib: bars and a legend

Code: python

import matplotlib.pyplot as plt

court_one = [4, 9, 6, 3]   # bookings at 9, 10, 11 and 12
court_two = [2, 5, 4, 1]

left = [8.8, 9.8, 10.8, 11.8]   # Court 1 bars a little left of each hour
right = [9.2, 10.2, 11.2, 12.2] # Court 2 bars a little right
plt.bar(left, court_one, width=0.4, label="Court 1")
plt.bar(right, court_two, width=0.4, label="Court 2")
plt.title("Bookings in each hour")
plt.xlabel("Hour of the day")
plt.ylabel("Number of bookings")
plt.legend()
plt.show()

A good chart has...

Part	Why
A title	Says what the chart is about
Labels on both axes	Says what the numbers mean, with units
The right kind of chart	Bars to compare, lines for change over time
A legend if there is more than one line or colour	Says what each one means
Clear colours	Anyone can read it, including people who cannot tell some colours apart

The cells students fill in are left blank in this preview.