Reading Python Objects and Records: Exercises

Q1. Distinguish three kinds of access

Which statement about these expressions is correct?

Choose one

path.name selects an attribute, text.split selects a method without calling it, text.split() calls that method, and record["name"] looks up a dictionary key.

All four expressions call functions because they select named values.

Dot notation and dictionary brackets are interchangeable forms of the same lookup.

Select one choice, then check.

HintLook for the call marker

Parentheses invoke the selected callable. Brackets use a value as a key.

SolutionRead each operator literally

The first expression selects data, the second selects a method object, the third calls it, and the fourth performs keyed lookup.

Q2. Use an attribute failure as evidence

This program fails:

What is the best first response?

Choose one

Inspect type(path) and the documented Path attributes, then use the supported attribute path.name if it matches the intended question.

Replace the expression with path["filename"] because every object also stores dictionary keys.

Catch every AttributeError and print an empty filename.

Select one choice, then check.

HintInspect the receiver

The exception says the selected object does not provide the requested attribute. First establish what that object is.

SolutionUse the supported Path attribute

Confirm that path is a Path, then use path.name for the final path component. Brackets and a broad handler would hide the real interface.

Q3. Map a signature to a call

Consider this documented learning signature:

Path.read_text(encoding=None, errors=None, newline=None) -> str

In path.read_text(encoding="utf-8"), which parameter uses an explicit value, and what happens to the other two parameters?

Choose one

encoding receives "utf-8"; errors and newline use their documented defaults.

All three parameters receive "utf-8" because one keyword value fills every default.

The call is incomplete because every parameter shown in a signature is required.

Select one choice, then check.

HintFollow the keyword name

A keyword argument states its parameter mapping directly.

SolutionUse one explicit option

encoding receives "utf-8". The call omits errors and newline, so both retain None for this documented signature.

Q4. Choose an explicit library call

The documented learning signature is numpy.linspace(start, stop, num=50). We need exactly five evenly spaced values from 0.0 through 1.0, including both endpoints. Which call states that choice most clearly?

Choose one

np.linspace(0.0, 1.0, num=5)

np.linspace(0.0, 1.0) because the default already means five values.

np.linspace(num=0.0, start=1.0, stop=5)

Select one choice, then check.

HintOverride only the changed default

The interval endpoints are required. Name the optional count because five is a deliberate departure from its default.

SolutionName the sample count

Use np.linspace(0.0, 1.0, num=5). The result contains five evenly spaced values, while the keyword makes the chosen count visible.

Q5. Trace instances and an alias

Predict the four printed values.

Four values, separated by commas

Answer it first, then check.

HintDraw two objects

same_celsius is another name for celsius. The fahrenheit constructor call created a different instance.

SolutionKeep identity separate from equality

The output is 21.4, 70.3, °C, and °F. Mutation through the alias is visible through celsius but does not reach the other instance.

Q6. Trace self through one call

Complete report_line so it reads fields from the instance that receives the call. The two calls must produce different unit labels.

Editable Python

HintRead through self

Inside the method, self is celsius for the first call and fahrenheit for the second.

SolutionReturn fields from the receiver

Q7. Interpret a hinted boundary

What does this signature state?

Choose one

The expected input is a list of floats or None markers, and the result may be a float or None; ordinary Python still does not enforce or convert those values automatically.

Python converts every list element to float before the function body runs.

The hint proves that the input has exactly six elements arranged along one named axis.

Select one choice, then check.

HintRead the bars as alternatives

float | None permits either named value form. It does not add executable checks to the function boundary.

SolutionTreat hints as interface notation

The parameter and return annotations describe expected alternatives. The body must still implement conversion, validation, and the all-unavailable result rule.

Q8. Read a frozen dataclass honestly

Which statement about this declaration is accurate?

Choose one

The dataclass generates ordinary record setup and blocks normal field reassignment, but it does not validate types or promise deep immutability for objects stored in fields.

Construction rejects every runtime value whose type differs from its annotation.

Every object reachable from an instance becomes recursively immutable.

Select one choice, then check.

HintKeep three claims separate

Ask what the dataclass generates, what frozen=True blocks, and what no runtime validator was asked to do.

SolutionUse the narrow frozen guarantee

Initialization, display, and equality are generated. Ordinary field reassignment is blocked. Type validation and recursive immutability are not supplied by this declaration.

Q9. Separate configuration, input, and result

Which assignment of roles is correct for a measurement run?

Choose one

Configuration: unit and thresholds. Input: measured values. Result: accepted/rejected counts, mean, and classification counts.

Configuration: unit, thresholds, every measured value, and every computed count.

Input: thresholds stored in global variables. Result: the original unprocessed file path only.

Select one choice, then check.

HintUse time and ownership

Configuration is chosen before analysis. Input is observed data. Result is evidence computed from that data under the chosen settings.

SolutionKeep the run boundary explicit

Unit and thresholds belong to MeasurementConfig; the reading list remains a separate input; counts and mean belong to MeasurementResult.

Q10. Complete the configuration-to-result flow

Complete analyze. Reject reversed or equal thresholds, treat None as one rejected input, classify accepted readings with the supplied configuration, and return a MeasurementResult. The same readings are analyzed with two configurations so the changed setting remains visible.

Editable Python

HintKeep one pass and three counters

After the threshold check, skip each None only after incrementing rejected. For a number, append it to accepted and update exactly one class counter.

SolutionReturn evidence under the chosen settings

Q1. Distinguish three kinds of access

Q2. Use an attribute failure as evidence

Q3. Map a signature to a call

Q4. Choose an explicit library call

Q5. Trace instances and an alias

Q6. Trace self through one call

Editable Python

Q7. Interpret a hinted boundary

Q8. Read a frozen dataclass honestly

Q9. Separate configuration, input, and result

Q10. Complete the configuration-to-result flow

Editable Python

Pause and reflect