Transform Sensor Readings - Array Processing Coding Interview

Question

## Problem

You receive a list of raw sensor readings as integers. Apply the following transforms in sequence:
1. **Normalize:** subtract the minimum value from every reading.
2. **Smooth:** replace each reading with the average of itself and its two neighbors (for endpoints, use only available neighbors). Round down to integer.
3. **Threshold:** set any reading below `t` to 0; leave others unchanged.
4. **Run-length encode:** return a list of `(value, count)` pairs for consecutive identical values.

```python
def transform_readings(readings: list[int], t: int) -> list[tuple[int, int]]:
    ...
```

**Example:**
```
readings = [5, 3, 8, 6, 4], t = 4
Step 1 (normalize): [2, 0, 5, 3, 1]
Step 2 (smooth):    [1, 2, 2, 3, 2]  # e.g. index 0: avg(2,0)=1, index 1: avg(2,0,5)=2
Step 3 (threshold): [0, 0, 0, 3, 0]  # values < 4 set to 0; wait, t=4: only 3<4? no: 1<4, 2<4, 2<4, 3<4.
                    Actually: [0,0,0,0,0] if all smoothed values < 4.
Output: [(0, 5)]
```

## Follow-ups
- What is the time complexity of each transform?
- How would you make this pipeline composable/configurable (add/remove steps at runtime)?
- Extend: step 2 uses a configurable window size `w` instead of fixed neighbors.

Transform Sensor Readings - Array Processing Coding Interview

Question Details

Full Details

Topics

More from Codesignal