Week 11 / Build a checkable report
Make the information a calculation depends on visible.
We have a reusable conversion. Reuse becomes unreliable if the function silently uses whichever sensor settings were changed most recently.
Sensor A has offset 0.5 V and sensor B has offset 1.0 V. Both use gain 25 N/V. Each reports 2.5 V.
Start with the problem
Calculate each force on paper. Should processing B first change the result for A? Explain what information each calculation needs.
Parameters pass configuration explicitly. Local variables keep each call’s work separate, so a small library can serve several sensors without hidden changes.
By the end: Process A, then B, then A again and explain why A’s two results agree.
A function that reads a global offset can change behaviour without changing its visible arguments. To reproduce the result, a reviewer then needs the history of previous cells as well as the current call.
Use a small record with named offset and gain fields. Pass that record to the conversion function. A dictionary makes the dependency visible, but it is still mutable: the team must decide who may modify it.
A conversion returns a numeric force. A report function can format it. A validation function can return a decision and a reason. Agree on record keys and tuple order so each part knows what the previous part produced.
Read the example alongside the explanation. Run it in a new notebook cell and change one input to see how it behaves.
def convert(voltage_v, calibration):
return (voltage_v - calibration['offset_v']) * calibration['gain_n_per_v']
sensor_a = {'offset_v': 0.5, 'gain_n_per_v': 25}
sensor_b = {'offset_v': 1.0, 'gain_n_per_v': 25}
print('A:', convert(2.5, sensor_a))
print('B:', convert(2.5, sensor_b))
print('A again:', convert(2.5, sensor_a))A: 50.0 B: 37.5 A again: 50.0
A local name can be rebound without rebinding the caller’s name. But if a parameter refers to the caller’s list, mutating that shared object can affect the caller. “Local variable” does not mean “private copy of every object.”
Draw caller and function frames, each with a name pointing to one list [2, 4]. Compare adding an element to that object with rebinding the local name to a new list.
Predict before running. Which changes the caller’s list: values.append(6), or values = [6] inside the function?
The first operation changes shared data; the second changes a local binding. Document intended mutation and pass sensor-specific settings as arguments when composing a reusable library.
Change one thing. Use the same helper with two sensor datasets in different orders. Predict which hidden global state could make the result depend on call order.
Türkçe: Kapsam adın nerede bulunduğunu belirler; mutasyon nesneyi değiştirir. Yerel parametre çağıranın nesnesini paylaşabilir.
Each example changes something about the same problem. Open the ones you want to explore and follow the worked explanation.
2.5 V; A offset 0.5 V; B offset 1.0 V; gain 25 N/V
Question: Predict the results of calls A, B, A.
The call names the calibration it needs. The order does not change the result while records stay unchanged.
A global offset changes from 0.5 to 1.0 V between identical convert(2.5) calls
Question: What hidden dependency changes the answer?
Reproducibility requires making the calibration and its version visible.
sensor_b = sensor_a; then sensor_b['offset_v'] = 1.0
Question: Has A's configuration changed too?
Explicit arguments help, but they do not prevent aliasing. Create independent records and avoid surprise mutation.
Interactive walkthroughs of Scope & Mini-Library. Enable JavaScript to step through code, variables, collections and output. The companion notebook remains available below.
Use the notebook to try the ideas yourself. The steps below connect this week's example to the programming practice.
Something to take away: A documented function interface and an A–B–A check showing that one sensor's configuration does not contaminate another's.
Read the notebook's teaching cells before these exercises.
Use the notebook's core and optional labels to choose your workload. This activity fits within guided class time.
Notes stay in this browser. Download a copy to keep them.
Ask AI to refactor a conversion library. Inspect every global read and every mutation of caller-owned records. Test the call order and verify that a claimed pure conversion does not change its inputs.
You can also review the supplied example with a partner. Use the same inputs to compare the reasoning. Follow the syllabus rules for assessed work.