DATA ANALYST FOUNDATIONS — MODULE 06
A guided portfolio project. Edition 2026-10-15.6.
Fictional Northstar Shop data and guided materials from DataCareerHub.

START
1 Extract portfolio_pack.zip into a local folder, retaining relative filenames.
2 Read project_brief.txt and use learner_worksheet.txt to attempt each milestone.
3 Preserve orders_raw.csv. Prepare your own copy, using orders_verified.csv as
   correction authority. Recreate your log before reading quality_log_solution.csv.
4 Choose a spreadsheet (Module 2 methods) or local SQLite (Module 3 methods).
   For SQLite, run analysis_setup.sql only in a NEW EMPTY LOCAL practice database.
   It creates raw_orders and a separately supplied reference orders table. It is
   not an automatic cleaning pipeline. Never run it against your hosting DB.
5 Complete analysis_starters.sql, or equivalent spreadsheet work. SQL window
   query P6 needs SQLite 3.25+. No paid software or account is required to read.
6 Compare analysis_solutions.sql and expected_outputs.json after attempting tasks.
7 Build your own dashboard, then compare dashboard_reference.html from Module 5.
8 Draft your own case study using project_starter.html or your preferred document.
   worked_portfolio.html shows a complete guided submission with local evidence links.
9 Add one independent extension; complete review_rubric.txt; inspect your sharing copy.

SOURCE / FIELD DICTIONARY
orders_raw.csv and orders_verified.csv are unchanged copies of Module 4 data.
Raw:46 records/44 IDs. Verified:44 unique orders. Completed:40 (20 each month).
order_id: stable fictional order key, not a customer. A=April, Y=May.
order_date: ISO YYYY-MM-DD. Reporting [2026-04-01,2026-06-01). No event timestamp.
channel: web/store after preparation; trim/lowercase is an authorised format rule.
status: completed/returned/cancelled. Not a fixed-follow-up customer return cohort.
order_value_eur: non-negative EUR MAJOR UNITS, not cents, profit or cash receipts.
delivery_days: non-negative integer measurement. Blank means not supplied, not zero.
Delivery comparison uses completed orders; measured mean excludes missing and
reports measured/eligible. A known numeric zero would remain measured.
Observed order dates end on each month's 22nd. Complete supplied fixture does not
independently prove every real-world transaction in the month was exported.

FILES
project_brief.txt, metric_contract.txt, learner_worksheet.txt, review_rubric.txt.
Raw/reference CSVs and completed quality_log_solution.csv.
analysis_setup.sql, analysis_starters.sql, analysis_solutions.sql, expected_outputs.json.
project_starter.html: editable layout prompts, not a completed or saved submission.
worked_portfolio.html: guided case study, with linked evidence in the extracted pack.
dashboard_reference.html: unchanged Module 5 interactive worked example, embedded
fictional data; no network, accounts or storage. A neighbouring CSV does not
refresh its data. Its Module 5 attribution is intentionally retained.
worked_solutions.txt: all milestone solutions, decision note and extension example.
portfolio_pack.zip: the 16 files above including this README, not a deployment ZIP.

METHOD / EXPECTATIONS
Read metric_contract.txt. SQL shares/coverage are fractions; change_pct is in
percentage units. P2 is a fixture-specific diagnostic, not a universal parser.
P6 compares previous observed rows; check missing calendar periods on new sources.
Keep baseline vs exclusion/equal-mix scenarios distinct. Verified A020=1000 stays
baseline. No costs, margins, customer exposure or experiment supplied; no profit,
ROI, customer conversion, stable forecast or causal effect established.
Work locally. No submission endpoint, registration, payment, certificate or
external publication. Attribute guided data/templates and describe your own work.

REFERENCES / CORRECTIONS
SQLite SELECT: https://www.sqlite.org/lang_select.html
SQLite windows: https://sqlite.org/windowfunctions.html
Earlier modules explain quality, exploration and selected usability references.
Case and exercises authored for DataCareerHub. Corrections: support@datacareerhub.io.
