How should I practice Detect Recurring Transactions with Tunable Tolerances?

Start by reading the full statement, solve the visible examples, then compare the pattern against related Ramp online assessment problems in FastPrep.

Is Detect Recurring Transactions with Tunable Tolerances associated with Ramp?

FastPrep tags this as a Ramp coding interview or online assessment practice problem when the problem catalog has that company context.

Problem · Hash Table

Detect Recurring Transactions with Tunable Tolerances

Learn this problem

● MediumRampFULLTIMEONSITE INTERVIEW

Problem statement

You are given an array transactions. Each record has the format merchant|YYYY-MM-DD|amountCents, where merchant is a non-empty case-sensitive string, the date is a valid ISO calendar date, and amountCents is a positive integer.

Return the merchants whose transactions form recurring subscriptions under the following rules:

Records are grouped only when their merchant strings are exactly equal.
A group must contain at least minOccurrences transactions.
The group's price spread, maximum amount - minimum amount, must be at most maxAmountDifferenceCents.
Sort the group's transactions by date and compute every consecutive gap in calendar days. Let the nominal interval be the lower median gap: after sorting the gaps, use the element at index (gaps.length - 1) / 2.
Every consecutive gap must differ from the nominal interval by at most maxIntervalDifferenceDays.

Return each qualifying merchant exactly once in ascending lexicographic order.

Function

findRecurringTransactions(transactions: String[], maxIntervalDifferenceDays: int, maxAmountDifferenceCents: int, minOccurrences: int) → String[]

Examples

Example 1

transactions = ["Netflix|2026-01-01|1599","Gym|2026-01-05|5000","Netflix|2026-01-31|1600","Gym|2026-01-12|5000","Netflix|2026-03-02|1598","Gym|2026-01-19|5000"]maxIntervalDifferenceDays = 1maxAmountDifferenceCents = 5minOccurrences = 3return = ["Gym","Netflix"]

Gym has two seven-day gaps and no price variation. Netflix has two thirty-day gaps and a price spread of 2 cents. Both satisfy the supplied thresholds, so they are returned in lexicographic order.

Example 2

transactions = ["Cloud|2026-01-01|1000","Cloud|2026-02-01|1300","Cloud|2026-03-01|1000","News|2026-01-10|999","News|2026-02-09|1025","News|2026-03-11|1000"]maxIntervalDifferenceDays = 2maxAmountDifferenceCents = 50minOccurrences = 3return = ["News"]

Cloud fails because its price spread is 300 cents. News has a 26-cent spread and equal thirty-day gaps, so it qualifies.

Example 3

transactions = ["Music|2026-01-01|1200","Music|2026-01-31|1200","Music|2026-04-01|1200","Storage|2026-01-15|500","Storage|2026-02-14|500"]maxIntervalDifferenceDays = 3maxAmountDifferenceCents = 0minOccurrences = 3return = []

Music has gaps of 30 and 60 days, so one gap is outside the three-day tolerance around the lower-median interval. Storage has fewer than three occurrences.

Constraints

1 <= transactions.length <= 200000
Every record has the exact format merchant|YYYY-MM-DD|amountCents.
Every date is a valid ISO calendar date from 1970-01-01 through 9999-12-31.
1 <= amountCents <= 10^9
0 <= maxIntervalDifferenceDays <= 3650
0 <= maxAmountDifferenceCents <= 10^9
2 <= minOccurrences <= transactions.length