Counters, Accumulators, Results, and Flags

Unit ID: M04-U06 Estimated active time: 25-35 minutes

Loops often build state

Four common patterns are:

• counter: how many items met a rule;
• accumulator: a running total;
• result collection: accepted, rejected, or transformed items;
• flag: whether an event occurred.

Count and total valid values

values = [8, 12, -2, 4]
valid_count = 0
valid_total = 0

for value in values:
    if value < 0:
        continue
    valid_count = valid_count + 1
    valid_total = valid_total + value

Initial values must represent an empty result: zero count and zero total.

Build accepted and rejected outputs

accepted = []
rejected = []

for value in values:
    if value < 0:
        rejected.append(value)
    else:
        accepted.append(value)

The outputs preserve the decision evidence.

Track whether anything failed

has_invalid_value = False

for value in values:
    if value < 0:
        has_invalid_value = True

If you only need to know whether any failure exists, you may break after setting the flag. If you need all rejection details, continue processing.

Avoid stale state

In a notebook, rerunning only the loop cell without resetting counters can double the totals. Keep initialisation in the same cell or ensure restart-and-run-all recreates clean state.

Use descriptive names

total is vague when several totals exist. Prefer accepted_hours_total, rejected_count, or found_target.

Practice

Process [8, 0, 41, "6", 12]. Build:

• valid integer values from 0 to 40;
• rejected values;
• accepted count;
• rejected count;
• accepted total; and
• a flag showing whether any rejection occurred.

Expected accepted values are [8, 0, 12] and accepted total is 20.

Takeaway

Initialise state before the loop, update it only in the correct branch, and keep names specific. Next, we will compare an explicit loop with a concise comprehension.