Power BI Fundamentals | MinAI Learning

🎯 Mục tiêu bài học

TB5 min

Sau bài học này, bạn sẽ:

✅ Hiểu Power BI ecosystem: Desktop, Service, Mobile

✅ Connect và transform data với Power Query

✅ Build star schema data model với relationships

✅ Viết DAX formulas: SUM, CALCULATE, Time Intelligence

✅ Tạo interactive reports với slicers và drill-through

✅ So sánh Power BI vs Tableau — khi nào dùng cái nào

Thời gian: 45 phút | Độ khó: Beginner | Tool: Power BI Desktop (Free)

Task 0

📖 Bảng Thuật Ngữ Quan Trọng

TB5 min

Thuật ngữ	Tiếng Việt	Mô tả
DAX	Data Analysis Expressions	Formula language của Power BI
Measure	Phép đo	Aggregation calculated at runtime
Calculated Column	Cột tính toán	Row-by-row calculation, stored in model
Star Schema	Mô hình sao	Fact table ở giữa, dimensions xung quanh
Power Query	Truy vấn nguồn	ETL tool tích hợp trong Power BI
Slicer	Bộ lọc trực quan	Interactive filter control trên report
Drill-through	Khoan sâu	Click → navigate to detail page
Bookmark	Đánh dấu	Lưu state cụ thể (filters, visible visuals)
Filter Context	Ngữ cảnh bộ lọc	Tập hợp filters đang active cho calculation
CALCULATE	Hàm CALCULATE	DAX function thay đổi filter context

Checkpoint

DAX = formula language của Power BI. Measure = aggregation at runtime. CALCULATE = thay đổi filter context. Star Schema = fact + dimensions. Bạn phân biệt Measure vs Calculated Column chưa?

Task 1

🖥️ 1. Power BI Ecosystem

TB5 min

1.1 Power BI Products

Product	Platform	Purpose	Cost
Power BI Desktop	Windows	Create reports	Free
Power BI Service	Web	Share & collaborate	Free/$10/mo
Power BI Mobile	iOS/Android	View on mobile	Free
Power BI Report Server	On-premise	Internal hosting	License

1.2 Power BI Desktop Interface

Power BI Desktop Layout

Khu vực	Vị trí	Nội dung
Ribbon	Top	File, Home, Insert, Modeling, View, Optimize, Help
Filters Pane	Left	Page/Visual/Drillthrough filters
Report Canvas	Center	Drag visuals here
Visualizations	Right-top	Chart type selector (▤▥▦▧▨▩)
Fields	Right-bottom	Tables and columns hierarchy
View Tabs	Bottom	📊 Report · 📋 Data · 🔗 Model

1.3 Three Views

View	Icon	Purpose
Report	📊	Build visualizations
Data	📋	View/edit tables
Model	🔗	Manage relationships

Power BI Desktop hoàn toàn miễn phí! Chỉ cần trả phí khi publish lên Service cho team collaboration. Download tại powerbi.microsoft.com

Checkpoint

Power BI có 3 views: Report (build visuals), Data (xem tables), Model (manage relationships). Desktop = Free, chỉ Service cần trả phí!

Task 2

📥 2. Getting Data & Power Query

TB5 min

2.1 Data Sources

📥Get Data → Categories

📄File (Excel, CSV, JSON, PDF)

🗄️Database (SQL Server, MySQL, PostgreSQL)

⚡Power Platform (Dataverse, Dataflows)

☁️Azure (Synapse, Blob Storage)

🌐Online Services (SharePoint, Dynamics 365)

🔧Other (Web, ODBC, Python/R scripts)

2.2 Load vs Transform

Option	When to Use
Load	Data is clean, ready to use
Transform	Need cleaning, reshaping

2.3 Power Query Editor

Power Query Editor Layout

Khu vực	Vị trí	Nội dung
Ribbon	Top	Home, Transform, Add Column, View, Tools, Help
Queries	Left-top	Orders, Products, Customers
Applied Steps	Left-bottom	Source → Navigation → Changed Type
Preview	Center	Data table preview (OrderID, Date, Customer, Amount)
Formula Bar	Center-top	`= Table.TransformColumnTypes(...)`

2.4 Common Transformations

Action	Use Case
Remove columns	Delete unnecessary fields
Filter rows	Remove nulls, errors
Change type	Text → Number, Date
Replace values	Clean up data
Split column	"John Doe" → "John", "Doe"
Merge queries	Join tables
Append queries	Stack tables
Pivot/Unpivot	Reshape data structure

Power Query = ETL trong Power BI. Mọi bước transform được ghi lại thành Applied Steps — có thể undo, reorder, hoặc edit bất cứ lúc nào!

Checkpoint

Power Query ghi lại mọi transform thành Applied Steps. Common transforms: remove columns, filter rows, change types, pivot/unpivot. Bạn đã dùng Transform Data chưa?

Task 3

🔗 3. Data Modeling

TB5 min

3.1 Star Schema

⭐FACT_Sales (SalesKey, DateKey, ProductKey, CustomerKey, Amount, Quantity)

📅DIM_Date

📋DateKey, Date, Month, Year

📦DIM_Product

📋ProductKey, ProductName, Category

👥DIM_Customer

📋CustomerKey, CustomerName, Segment

3.2 Creating Relationships

Model View → Drag field from Table A to Table B

🔗Relationship Properties

🔢Cardinality: One-to-Many, Many-to-Many

🔀Cross Filter: Single, Both

✅Active: Yes/No

⚡Referential Integrity: Assume (for performance)

3.3 Best Practices

✅ DO:

Use star schema (facts + dimensions)
Hide technical keys from report view
Create date dimension table
Name tables and columns clearly

❌ DON'T:

Many-to-many without bridge table
Circular dependencies
Too many bi-directional filters
Wide tables với 100+ columns

Star Schema là foundation của mọi Power BI model tốt. Fact table chứa numbers (Sales, Quantity), Dimension tables chứa attributes (Product Name, Region). Không có star schema → DAX sẽ rất khó viết!

Checkpoint

Star Schema: Fact table (numbers) ở giữa, Dimension tables (attributes) xung quanh. Relationships thường One-to-Many, single direction. Bạn đã tạo model view chưa?

Task 4

📐 4. DAX Basics

TB5 min

4.1 Calculated Column vs Measure

Aspect	Calculated Column	Measure
Calculated	Row by row	On aggregation
Storage	Stored in model	Calculated at runtime
Use	Categorization, filtering	Aggregations, KPIs
Performance	Uses memory	More efficient

4.2 Essential DAX Functions

dax

1-- SUM, AVERAGE, COUNT
2Total Sales = SUM(Sales[Amount])
3Avg Order = AVERAGE(Sales[Amount])
4Order Count = COUNT(Sales[OrderID])
5 
6-- DISTINCTCOUNT
7Customer Count = DISTINCTCOUNT(Sales[CustomerID])
8 
9-- CALCULATE - thay đổi filter context
10Sales East = CALCULATE(
11    SUM(Sales[Amount]),
12    Region[RegionName] = "East"
13)
14 
15-- Time Intelligence
16Sales YTD = TOTALYTD(SUM(Sales[Amount]), 'Date'[Date])
17Sales PY = CALCULATE(
18    SUM(Sales[Amount]),
19    SAMEPERIODLASTYEAR('Date'[Date])
20)
21 
22-- Year-over-Year Growth
23YoY Growth = 
24VAR CurrentYear = SUM(Sales[Amount])
25VAR LastYear = CALCULATE(
26    SUM(Sales[Amount]),
27    SAMEPERIODLASTYEAR('Date'[Date])
28)
29RETURN 
30DIVIDE(CurrentYear - LastYear, LastYear, 0)

4.3 Filter Context & ALL

dax

1-- ALL removes filter context
2Total All Products = CALCULATE(
3    SUM(Sales[Amount]),
4    ALL(Product)
5)
6 
7-- % of Total
8% of Total = 
9DIVIDE(
10    SUM(Sales[Amount]),
11    CALCULATE(SUM(Sales[Amount]), ALL(Sales)),
12    0
13)

4.4 Variables (VAR)

dax

1Profit Margin = 
2VAR TotalRevenue = SUM(Sales[Revenue])
3VAR TotalCost = SUM(Sales[Cost])
4VAR Profit = TotalRevenue - TotalCost
5RETURN
6DIVIDE(Profit, TotalRevenue, 0)

CALCULATE là hàm quan trọng nhất trong DAX — nó thay đổi filter context. VAR giúp code clean và readable. Luôn dùng DIVIDE thay vì / để tránh divide-by-zero errors!

Checkpoint

CALCULATE = thay đổi filter context. ALL = remove filters. DIVIDE = safe division. VAR = clean code. Measure > Calculated Column cho performance!

Task 5

📊 5. Building Reports

TB5 min

5.1 Visual Types

Visualizations Pane

Icon	Visual Type
▤	Stacked bar
▥	Clustered bar
▦	Stacked column
▧	Line chart
▨	Area chart
▩	Combo chart
○	Pie/Donut
⬡	Treemap
🗺	Map
📊	Matrix
📋	Table
📇	Card
🎯	KPI
⬚	Slicer

5.2 Card Visuals for KPIs

KPI Cards

KPI	Value	Change
Total Sales	$2.5M	▲ 15%
Profit	$420K	▲ 8%
Orders	12,547	▲ 22%

5.3 Slicers & Filters

Slicer types:

List (checkboxes)
Dropdown
Date range (slider)
Relative date ("Last 30 days")

Checkpoint

Power BI có 15+ visual types. Card cho KPIs, Slicer cho interactive filters. Drag fields vào Axis, Values, Legend, Tooltips wells. Bạn đã tạo report đầu tiên chưa?

Task 6

🔄 6. Interactivity & Advanced

TB5 min

6.1 Cross-filtering

Ví dụ

1Click bar in Chart A → 
2  Chart B filters to show related data
3  Chart C highlights related data

6.2 Edit Interactions

Format → Edit Interactions

🖱️For each visual, choose

🔍Filter (subset data)

✨Highlight (dim non-related)

⏹️None (no interaction)

6.3 Drill-through

Ví dụ

1Setup:
21. Create detail page
32. Add drill-through field (e.g., Product)
43. Right-click on visual → Drill through
5 
6Main Report: Click "Electronics" →
7  Drill through → Detail page filtered to Electronics

6.4 Bookmarks

Save specific states: Filters applied, Visuals visible/hidden, Page navigation

Use for: Toggle views (detailed/summary), Guided navigation, Presentations

Checkpoint

Cross-filtering = click 1 chart → filter others. Edit Interactions = control filter/highlight/none. Drill-through = navigate to detail page. Bookmarks = saved states!

Task 7

⚖️ 7. Power BI vs Tableau

TB5 min

Comparison

Aspect	Power BI	Tableau
Price	$10/user/mo	$70/user/mo
Ease	Easier start	Steeper curve
Visualization	Good	Excellent
Data Modeling	Excellent	Good
DAX/Calcs	Very powerful	Powerful
Microsoft	Native integration	Connectors
Community	Growing	Mature

When to Use Which?

Choose Power BI:

Microsoft 365 organization
Budget-conscious
Need strong data modeling
Self-service BI focus

Choose Tableau:

Visualization-heavy needs
Complex analytics
Cross-platform teams
Premium features needed

Cả hai đều tuyệt vời — Power BI mạnh về data modeling + Microsoft integration, Tableau mạnh về visualization + flexibility. Biết cả hai = competitive advantage lớn!

Checkpoint

Power BI: rẻ hơn, Microsoft integration, strong modeling. Tableau: visualization premium, cross-platform. Biết cả hai là competitive advantage trong job market!

Task 8

🏋️ 8. Hands-on: Sales Report

TB5 min

Project

Build executive sales dashboard hoàn chỉnh trong Power BI!

Page 1: Executive Summary

Executive Summary Layout

Vị trí	Nội dung
KPI Row	💰 Revenue: $2.5M · 📈 Growth: +15% · 📦 Orders: 12K · 👥 Customers: 3.2K
Middle Left	Revenue by Month (Line)
Middle Right	Revenue by Region (Map)
Bottom Left	Top Products (Bar)
Bottom Right	Segment Split (Donut)
Filters	Date Slicer · Region · Segment

Page 2: Product Detail (Drill-through)

Key DAX Measures

dax

1Revenue = SUM(Sales[Amount])
2Revenue YTD = TOTALYTD([Revenue], 'Date'[Date])
3Revenue PY = CALCULATE([Revenue], SAMEPERIODLASTYEAR('Date'[Date]))
4Growth % = DIVIDE([Revenue] - [Revenue PY], [Revenue PY], 0)
5Profit Margin = DIVIDE(SUM(Sales[Profit]), SUM(Sales[Revenue]), 0)

Task 9

📋 Tổng kết

TB5 min

Kiến thức đã học

Chủ đề	Nội dung chính	Tầm quan trọng
Ecosystem	Desktop (free), Service, Mobile	Overview
Power Query	ETL, Applied Steps, transforms	Data prep
Data Model	Star schema, relationships	Foundation
DAX	CALCULATE, Time Intelligence, VAR	Core skill
Reports	Cards, Slicers, 15+ visual types	Building
Interactivity	Cross-filter, Drill-through, Bookmarks	Advanced
vs Tableau	Price, features, when to use	Career

Câu hỏi tự kiểm tra

Power Query vs DAX: khi nào dùng cái nào?
Star schema là gì và tại sao quan trọng?
CALCULATE trong DAX dùng để làm gì?
Cross-filter và cross-highlight khác nhau thế nào?

Bài tiếp theo: Design Best Practices — Nguyên tắc thiết kế dashboard đẹp và hiệu quả!

🎉 Tuyệt vời! Bạn đã nắm vững Power BI Desktop!

Nhớ: Power Query để clean data, DAX để tính toán, Report để trình bày — 3 bước cho mọi dự án BI!

Task 10