DATA1002 Informatics: Data and Computation
The Complete Exam Bible · S2 2026DATA1002 Overview
- The University of Sydney
- Semester 2, 2026
- an undergraduate introductory data and computation unit
- 6 credit points
- a foundation in Python
DATA1002 Informatics: Data and Computation develops Python programming, data ingestion and cleaning, aggregation, visualisation, machine learning, communication, ethics and human-AI judgement. It is taught within The University of Sydney School of Computer Science. It is an undergraduate introductory data and computation unit.
- Two pass controls The exam requires 40% plus 50 overall; separately, attendance for at least 6 weeks across Week 4 to Week 13 is required unless Special Consideration applies, or an AF is automatic.
- Project arc Stage 1 documents data, cleaning, summaries and charts; Stage 2 presents a predictive model and evaluates its success.
- Two closed-book checks The 60-minute Python coding test and two-hour written exam verify different kinds of individual understanding.
- DATA1002 boundary Do not import DATA1902 weights: the shared teaching site contains two distinct assessment architectures.
How DATA1002 is assessed
| Component | Weight | Format |
|---|---|---|
| Written Examination · hurdle | 50% | Two-hour closed-book MCQ and short-essay exam; 40% minimum |
| Weekly Coding Tasks | 10% | Python tasks graded through automated output checks |
| Early-feedback MCQ | 2% | Week 3 check on Weeks 1-3 |
| Practice Coding Test | 0% | Unweighted preparation for the in-class coding test |
| Project Stage 1 Report | 20% | Group report on data, cleaning, summaries and charts |
| Python Coding Test | 10% | 60-minute closed-book in-class automated test |
| Project Stage 2 Poster | 8% | In-class group presentation on predictive modelling |
DATA1002 totals 100% across a 50% written exam, 10% weekly coding, 2% early MCQ, 20% project report, 10% coding test and 8% poster; the practice coding test is 0%. The written exam is the only hurdle row in the assessment table, with at least 40% on the exam and 50 overall required. Separately, attendance is required for at least 6 weeks from Week 4 to Week 13 inclusive; without approved Special Consideration, attending fewer than 6 weeks automatically results in an AF grade. DATA1902 weights are excluded.
Current DATA1002 dates
| Date | Item | Control |
|---|---|---|
| 23 August 2026 at 5:00 pm | Early-feedback MCQ | Covers Weeks 1-3 and contributes 2%. |
| 11 October 2026 at 5:00 pm | Project Stage 1 report | Group data report worth 20%. |
| Formal examination period | Written examination | No exact date is published in the outline. |
Dates are as published in the current course materials. Confirm exact deadlines and submission settings in the live LMS.
What DATA1002 covers
Read Data Questions, Lifecycle and Project Failure as the foundation, Data Visualisation and Communication as the main change in method, and Project Report, Poster and Secure Assessment as the final application of the course.
Data Questions, Lifecycle and Project Failure
data science lifecycle · decision question · project failure mode · turn a broad request into a decision question and map the data work needed to answer itDefine data science lifecycle
The course material gives this chapter a concrete anchor: The current introduction distinguishes data-science work from a linear software project and places iteration, evidence and decision use before modelling.
- data science lifecycle
- decision question
- project failure mode
- Applying data science lifecycle
- Limits of decision question and project failure mode
Python Values, Types and Testable Computation
Python value · data type · automated test · write a small calculation whose types, transformations and expected output can be checked independentlyDefine Python value
The course material gives this chapter a concrete anchor: The current Python introduction and first lab move from calculator expressions to types, built-in functions and observable output checks.
- Python value
- data type
- automated test
- Applying Python value
- Limits of data type and automated test
Conditionals, Loops and Text Processing
conditional branch · iteration · text parsing · convert a text-processing requirement into branches and loops with boundary testsDefine conditional branch
The course material gives this chapter a concrete anchor: The current Week 2 materials combine Python control flow with text processing and cleaning tasks rather than treating syntax as isolated recall.
- conditional branch
- iteration
- text parsing
- Applying conditional branch
- Limits of iteration and text parsing
Lists, Dictionaries and Aggregation Patterns
list · dictionary · aggregation · choose data structures and aggregate records without losing the grouping key or denominatorDefine list
The course material gives this chapter a concrete anchor: The official Weeks 4-5 sequence pairs lists and dictionaries with bucketing, pivoting and data-quality reasoning.
- list
- dictionary
- aggregation
- Applying list
- Limits of dictionary and aggregation
Data Cleaning, Provenance and Quality
data provenance · missing-value convention · data-quality rule · build a cleaning pipeline whose assumptions and row-level effects can be auditedDefine data provenance
The course material gives this chapter a concrete anchor: The lifecycle, lab and project materials require students to document source, metadata, cleaning and ingestion before reporting summaries.
- data provenance
- missing-value convention
- data-quality rule
- Applying data provenance
- Limits of missing-value convention and data-quality rule
Functions, Modules, CSV and Pandas
Python function · module · DataFrame · organise ingestion and transformation into reusable functions with explicit inputs and outputsDefine Python function
The course material gives this chapter a concrete anchor: The official Week 6 sequence introduces csv, pandas, modules and functions as the project shifts from exercises to reusable data work.
- Python function
- module
- DataFrame
- Applying Python function
- Limits of module and DataFrame
Data Visualisation and Communication
visual encoding · comparison baseline · chart integrity · select and design a chart that supports the intended comparison and communicates its evidence boundaryDefine visual encoding
The course material gives this chapter a concrete anchor: Week 8 explicitly combines chart concepts, chart evaluation, Matplotlib and trustworthy decision making for the Stage 1 report.
- visual encoding
- comparison baseline
- chart integrity
- Applying visual encoding
- Limits of comparison baseline and chart integrity
Clustering, Similarity and Recommendation
clustering · distance metric · recommendation · explain how representation and distance produce clusters or recommendations and evaluate whether the grouping is usefulDefine clustering
The course material gives this chapter a concrete anchor: The official Week 9 topic introduces machine learning through clustering and recommendation with scikit-learn practice.
- clustering
- distance metric
- recommendation
- Applying clustering
- Limits of distance metric and recommendation
Predictive Models, Evaluation and Explainability
predictive model · evaluation metric · generalisation · fit and evaluate a predictive model using held-out evidence and a metric aligned to the decisionDefine predictive model
The course material gives this chapter a concrete anchor: Weeks 10-11 and Project Stage 2 require a predictive model, evaluation of success and explanation through a group poster.
- predictive model
- evaluation metric
- generalisation
- Applying predictive model
- Limits of evaluation metric and generalisation
Natural Language Processing and Annotation
natural language processing · data annotation · inter-annotator agreement · design a language-labelling task and distinguish model error from an unclear category schemeDefine natural language processing
The course material gives this chapter a concrete anchor: Weeks 11-12 connect NLP, examples, data labelling and human disagreement as data-quality and modelling issues.
- natural language processing
- data annotation
- inter-annotator agreement
- Applying natural language processing
- Limits of data annotation and inter-annotator agreement
GenAI Reliability, Ethics and Fairness
generative model · algorithmic fairness · human oversight · audit an AI-assisted data decision for provenance, uncertainty, subgroup impact and accountable reviewDefine generative model
The course material gives this chapter a concrete anchor: The current course threads GenAI reasoning through early weeks and closes with ethics and fairness rather than isolating AI from the data lifecycle.
- generative model
- algorithmic fairness
- human oversight
- Applying generative model
- Limits of algorithmic fairness and human oversight
Project Report, Poster and Secure Assessment
reproducible analysis · poster narrative · exam hurdle · integrate coding, project evidence and exam reasoning while keeping DATA1002 facts separate from DATA1902Define reproducible analysis
The course material gives this chapter a concrete anchor: The official table and assessment page distinguish unweighted practice, individual coding, two project stages and the sole written-exam hurdle.
- reproducible analysis
- poster narrative
- exam hurdle
- Applying reproducible analysis
- Limits of poster narrative and exam hurdle
It carries 6 credit points. It is positioned as a foundation in Python, data handling, visualisation and predictive modelling.
The current sequence joins computational fundamentals to a staged group data project.
It explicitly separates DATA1002 from advanced DATA1902 and uses a closed-book Python test plus a closed-book final examination to verify individual capability.
Assessment in DATA1002 is distributed as follows: 50% written exam, 10% weekly coding, 2% early-feedback MCQ, 20% project report, 10% closed-book Python coding test and 8% project poster; a practice coding test carries 0%
The operational assessment conditions matter here.
The final is a two-hour closed-book examination with multiple-choice and short-essay questions during the formal examination period.
The Python coding test is a separate 60-minute closed-book in-class task with no cheat sheet.
What makes DATA1002 demanding is concrete: Turning messy data into a defensible decision: students must write correct Python, document transformations, choose meaningful summaries and charts, evaluate predictive models, and distinguish evidence from confident-looking AI or algorithmic output.
The written examination is the only hurdle row in the assessment table: students need at least 40% in that examination and at least 50 overall, with failure potentially capping the final mark at 45. A separate attendance pass rule requires at least 6 weeks from Week 4 to Week 13 inclusive;
without approved Special Consideration, fewer than 6 weeks automatically results in an AF grade.
For enrolment planning, The current outline lists no prerequisites or assumed knowledge and prohibits INFO1903 and DATA1902.
Read Data Questions, Lifecycle and Project Failure as the foundation, Data Visualisation and Communication as the main change in method, and Project Report, Poster and Secure Assessment as the final application of the course.
Coverage note: DATA1902 shares teaching pages but has a different assessment split;
this build uses only the DATA1002 rows.
Worked example · free
Diagnose a data-cleaning bug before fitting a model
- 1Inspect types, ranges and missing-value conventions before changing the data.
- 1Replace the documented age sentinel -1 with a missing value rather than a real age.
- 1Remove currency symbols and separators, then convert price to a numeric type with invalid conversions exposed.
- 1Record row counts and summary statistics before and after each transformation.
- 1Write automated checks that ages are plausible and nonmissing prices are numeric.
- 1Fit only after deciding how missing ages will be excluded, imputed or represented and documenting that decision.
Key terms
- Data cleaning
- The documented process of detecting and correcting invalid, inconsistent, missing or unsuitable data values.
- Predictive model
- A fitted rule that uses input features to estimate an unknown outcome for new observations.
- Algorithmic fairness
- The evaluation of how a computational system distributes errors, benefits and burdens across relevant groups.
- Data science lifecycle
- An iterative sequence for framing a question, obtaining data, cleaning, analysing, modelling, communicating and reviewing.
- Python function
- A named reusable block of code that accepts inputs, performs operations and may return a result.
- Aggregation
- The combination of multiple observations into grouped summaries such as counts, totals, means or rates.
- Pandas DataFrame
- A labelled two-dimensional tabular data structure whose columns can hold different data types.
- Data visualisation
- A graphical encoding of data designed to reveal structure and support accurate comparison or communication.
- Clustering
- An unsupervised method that groups observations according to a defined measure of similarity or distance.
- Natural language processing
- Computational methods for representing, analysing or generating human language data.
- Model evaluation
- The structured assessment of predictive performance using suitable data, metrics and comparison baselines.
- Data provenance
- Recorded information about where data came from and how they were collected, transformed and used.
- Exception handling
- Program logic that anticipates and responds to runtime errors without silently corrupting the result.
DATA1002 FAQ
Which current DATA1002 dates are captured?
Early-feedback MCQ: 23 August 2026 at 5:00 pm; Project Stage 1 report: 11 October 2026 at 5:00 pm; Written examination: Formal examination period. Confirm any change and the exact submission setting in the live LMS.
How is DATA1002 assessed?
50% written exam, 10% weekly coding, 2% early-feedback MCQ, 20% project report, 10% closed-book Python coding test and 8% project poster; a practice coding test carries 0%
What is the DATA1002 exam or final-task format?
The final is a two-hour closed-book examination with multiple-choice and short-essay questions during the formal examination period. The Python coding test is a separate 60-minute closed-book in-class task with no cheat sheet.
Does DATA1002 have a hurdle or component-level pass rule?
The written examination is the only hurdle row in the assessment table. School policy requires at least 40% in that examination and an overall final mark of at least 50; failure may cap the final mark at 45. A separate attendance pass rule requires at least 6 weeks from Week 4 to Week 13 inclusive; without approved Special Consideration, fewer than 6 weeks automatically results in an AF grade.
What prerequisites or restrictions apply to DATA1002?
The current outline lists no prerequisites or assumed knowledge and prohibits INFO1903 and DATA1902.
Where do students usually lose marks in DATA1002?
Turning messy data into a defensible decision: students must write correct Python, document transformations, choose meaningful summaries and charts, evaluate predictive models, and distinguish evidence from confident-looking AI or algorithmic output.
Which offering does this DATA1002 guide cover?
It is aligned to Semester 2, 2026; confirm your enrolled class and timetable in the current institutional system.
Is this DATA1002 resource an official university guide?
No. It is an independent DATA1002 study resource; current institutional instructions remain authoritative for assessment operation.
Questions students ask in each DATA1002 chapter
- What is the main task in Data Questions, Lifecycle and Project Failure?
- Turn a broad request into a decision question and map the data work needed to answer it. Chapter 01 · Data Questions, Lifecycle and Project Failure
- What is the main task in Python Values, Types and Testable Computation?
- Write a small calculation whose types, transformations and expected output can be checked independently. Chapter 02 · Python Values, Types and Testable Computation
- What is the main task in Conditionals, Loops and Text Processing?
- Convert a text-processing requirement into branches and loops with boundary tests. Chapter 03 · Conditionals, Loops and Text Processing
- What is the main task in Lists, Dictionaries and Aggregation Patterns?
- Choose data structures and aggregate records without losing the grouping key or denominator. Chapter 04 · Lists, Dictionaries and Aggregation Patterns
- What is the main task in Data Cleaning, Provenance and Quality?
- Build a cleaning pipeline whose assumptions and row-level effects can be audited. Chapter 05 · Data Cleaning, Provenance and Quality
- What is the main task in Functions, Modules, CSV and Pandas?
- Organise ingestion and transformation into reusable functions with explicit inputs and outputs. Chapter 06 · Functions, Modules, CSV and Pandas
- What is the main task in Data Visualisation and Communication?
- Select and design a chart that supports the intended comparison and communicates its evidence boundary. Chapter 07 · Data Visualisation and Communication
- What is the main task in Clustering, Similarity and Recommendation?
- Explain how representation and distance produce clusters or recommendations and evaluate whether the grouping is useful. Chapter 08 · Clustering, Similarity and Recommendation
- What is the main task in Predictive Models, Evaluation and Explainability?
- Fit and evaluate a predictive model using held-out evidence and a metric aligned to the decision. Chapter 09 · Predictive Models, Evaluation and Explainability
- What is the main task in Natural Language Processing and Annotation?
- Design a language-labelling task and distinguish model error from an unclear category scheme. Chapter 10 · Natural Language Processing and Annotation
- What is the main task in GenAI Reliability, Ethics and Fairness?
- Audit an ai-assisted data decision for provenance, uncertainty, subgroup impact and accountable review. Chapter 11 · GenAI Reliability, Ethics and Fairness
- What is the main task in Project Report, Poster and Secure Assessment?
- Integrate coding, project evidence and exam reasoning while keeping data1002 facts separate from data1902. Chapter 12 · Project Report, Poster and Secure Assessment
How to study for the exam
Retrieve the course map, practise the recurring method—frame the data question, inspect provenance and structure, write testable Python transformations, choose a visual or model suited to the decision, and evaluate the result for error, fairness and transfer—on changed scenarios, and verify every operational assessment detail in the live institutional system.
Your AI Data Science tutor for DATA1002
Stuck on a hard DATA1002 question? Sia is AskSia’s AI Data Science tutor — ask any DATA1002 Informatics: Data and Computation question and get a clear, step-by-step explanation grounded in how the course is actually taught and assessed. Read this whole study guide free, then take your hardest questions to Sia.
