Seconds to HH:MM — Solution & ExplanationProblems / Seconds to HH:MM / EditorialApproach

Hours: duration_seconds // 3600 — integer division gives whole hours.

Minutes: (duration_seconds % 3600) // 60 — the modulo strips the hours, then integer division gives whole minutes. Leftover seconds vanish in the division, which is exactly the "ignore leftover seconds" requirement.

Format: cast each part to VARCHAR, zero-pad with lpad(part, 2, '0'), and join with || ':' ||.

Key Insightlpad is a *string* function with a fixed output length — it pads short input but also truncates long input. That makes the cast discipline matter: in PySpark, F.floor returns a double, so casting it straight to string yields "1.0", and lpad("1.0", 2, "0") truncates it to "1.". Cast to an integer type first. The duration_seconds // 3600 / % 3600 // 60 decomposition is the same trick used for any unit conversion (bytes → GB/MB, cents → dollars/cents).Hints

Split the math first: hours = duration_seconds // 3600, minutes = (duration_seconds % 3600) // 60. Integer division drops the leftover seconds automatically.

lpad only works on strings, so CAST the numbers to VARCHAR first, then lpad(..., 2, '0') gives the zero-padding.

Glue the pieces with || ':' || in SQL, or F.concat(..., F.lit(":"), ...) in PySpark.

In PySpark, F.floor returns a double — cast to "int" before casting to "string", otherwise you'll lpad "1.0" and get garbage like '1.:23'.

Solve Seconds to HH:MM yourself →