Topic 4.2 Notes – Introduction to Using Data Sets

1. What a Data Set Is

A data set is a collection of related pieces of information.

“Related” means each value represents the same type of thing:

• Test scores from one class
• Daily step counts for a month
• Monthly sales numbers
• Attendance records (present/absent)

Each value follows the same structure:

• All numbers (like temperatures)
• All words (like product names)
• All boolean values (true/false)

The shift in thinking is this:

• We don’t care about one value by itself.
• We care about what the entire collection tells us.

Instead of writing separate variables like score1, score2, score3, you think:

“How do I process every score the same way?”

That idea of scalability matters a lot. A good algorithm works whether there are 5 values or 5 million. You design a pattern, not one-off code.

2. Sequential Processing One Value at a Time

When analyzing a data set, you usually access values one at a time.

You:

1. Look at one value
2. Update some tracking information
3. Move to the next value
4. Repeat

That pattern is the backbone of almost every array FRQ later.

The Core Processing Pattern

Most data set algorithms follow this structure:

1. Initialize tracking variables
2. Process each value in the set
3. Update tracking variables as needed
4. Use the final tracked result

For example, imagine tracking daily temperatures. As you move day by day, you update a running total and keep track of the highest temperature seen so far.

Stepwise temperature tracking with running total and max

Each row represents one step of the algorithm:

• The running total increases as each new temperature is added.
• The max only changes when a larger value appears.

You process one row at a time. That’s it.

Common Types of Data Set Operations

You’ll see the same patterns over and over.

a. Sum

• Start with sum = 0
• Add each value
• Final sum is total

b. Average

• Compute sum
• Divide by number of values
• Be careful with integer vs double division

Students lose points here by dividing too early or forgetting to use a double.

c. Count with a Condition

• Start count = 0
• If value meets condition → count++
• Final count is answer

Example idea: Count how many temperatures are below 32.

d. Maximum or Minimum

• Start with the first value as current max
• Compare each new value
• Replace if larger

Huge common mistake: starting max at 0.
That breaks if all numbers are negative.

e. Pattern Detection

Example: longest streak of consecutive absences.

You might track:

• currentStreak
• maxStreak

Each value updates those trackers.

This kind of logic shows up a lot on tougher FRQs.

3. Representing Data with Tables and Diagrams

Before coding, it helps to visualize the data.

Data can be represented as:

• Tables
• Charts
• Labeled diagrams

Here’s a simple example of a data table showing sales for a week:

Weekly sales data table

Looking at a table like this helps you:

• See patterns
• Identify what needs to be tracked
• Plan your algorithm

You can even add extra columns like:

• Running total
• Count
• Max so far

That visual planning step makes tracing much easier. On in-class tests or the AP exam, rewriting small data sets as tables on scratch paper can save you from logic mistakes.

4. Planning Before Coding

Strong problem-solvers don’t jump straight into syntax.

They think:

• What question am I answering?
• What do I need to track?
• What changes each time I see a new value?

If the question is “How many days were above 80 degrees?”
You only need:

• A counter
• A condition check

If the question is “What was the highest temperature?”
You need:

• A current max
• A comparison each step

Everything comes back to:

For each value, do the same process.

That mindset prepares you for:

• Arrays
• ArrayLists
• 2D arrays
• Most of FRQ 3 and FRQ 4

5. Edge Cases and Common Mistakes

Even at this conceptual stage, think about edge cases:

• Empty data set
• Only one value
• All values the same
• All negative numbers

Typical mistakes:

• Forgetting to initialize correctly
• Dividing before finishing counting
• Resetting a tracking variable at the wrong time
• Thinking about individual values instead of patterns

Data set problems almost always reduce to:

Process each value, track something, return the result.