Appearance
🛠 Workshop — Project Kickoff: Từ Idea đến Git Repo
Chọn đề tài → download dataset → data audit → viết project brief → setup Git repo + README → push lên GitHub. Output: Git repo hoàn chỉnh + project brief + data audit report — base cho Buổi 19 (Analysis & Dashboard)!
🎯 Mục tiêu workshop
Sau khi hoàn thành workshop này, bạn sẽ:
- Chọn đề tài + dataset — đã evaluate theo DRBST + checklist 10 điểm
- Viết project brief — 5 business questions + methods + tools
- Setup Git repo — folder structure chuẩn + .gitignore + README.md
- Data audit — profile dataset, đánh giá chất lượng, document findings
🧰 Yêu cầu
| Yêu cầu | Chi tiết |
|---|---|
| Kiến thức | Đã hoàn thành Buổi 18 lý thuyết (Project Kickoff) |
| Tools | Python (pandas, numpy), Git, GitHub account, text editor |
| Thời gian | 90–120 phút |
| Output | Git repo (GitHub) + Project Brief + Data Audit Report |
💡 Naming convention
Repo name: capstone-[domain]-analysis (ví dụ: capstone-ecommerce-analysis)
📦 Scenario
Bạn là fresh grad DA, chuẩn bị portfolio project đầu tiên. Mục tiêu: 2 tuần hoàn thành capstone project — từ data đến dashboard đến presentation. Hôm nay (Day 1): kickoff — chọn đề tài, setup repository, audit data.
Cuối workshop, bạn phải có:
- ✅ GitHub repo live (public)
- ✅ Project brief document
- ✅ Data audit report
- ✅ Clean folder structure
- ✅ README.md v1
Phần 1: Chọn Đề Tài + Dataset (20 phút)
Bước 1.1: Brainstorm & Evaluate Ideas
Dùng bảng DRBST để đánh giá 3-5 ideas:
python
# ============================================
# PHẦN 1: CHỌN ĐỀ TÀI — DRBST EVALUATION
# ============================================
import pandas as pd
# Brainstorm ideas — điền ideas của bạn vào đây
ideas = pd.DataFrame({
'Idea': [
'E-commerce Customer Segmentation (Olist)',
'HR Employee Attrition (IBM)',
'Marketing Campaign ROI',
'Telecom Customer Churn',
# 'Your idea here'
],
'D_Data': [5, 5, 4, 5], # 1-5: Data available + quality
'R_Real_Context': [5, 4, 4, 5], # 1-5: Real business context
'B_Business_Impact': [5, 5, 4, 4], # 1-5: Actionable insights
'S_Scope_2weeks': [4, 5, 4, 4], # 1-5: Feasible in 2 weeks
'T_Tellable': [5, 4, 4, 4] # 1-5: Can present in 5 min
})
ideas['Total'] = ideas[['D_Data', 'R_Real_Context', 'B_Business_Impact',
'S_Scope_2weeks', 'T_Tellable']].sum(axis=1)
ideas['Average'] = ideas['Total'] / 5
print("=" * 70)
print("📊 DRBST EVALUATION — CAPSTONE IDEAS")
print("=" * 70)
print(ideas.sort_values('Total', ascending=False).to_string(index=False))
print(f"\n🏆 Best Idea: {ideas.loc[ideas['Total'].idxmax(), 'Idea']}")
print(f" Score: {ideas['Total'].max()}/25 (Average: {ideas['Average'].max():.1f}/5)")Bước 1.2: Download Dataset
Sau khi chọn đề tài, download dataset. Ví dụ với Olist E-commerce (Kaggle):
python
# ============================================
# DOWNLOAD DATASET
# ============================================
# Option 1: Kaggle CLI (nếu đã setup Kaggle API)
# Terminal: kaggle datasets download -d olistbr/brazilian-ecommerce
# Option 2: Manual download từ Kaggle
# Link: https://www.kaggle.com/datasets/olistbr/brazilian-ecommerce
# Option 3: Dùng dataset gợi ý (smaller, ready-to-use)
# Tạo sample dataset nếu chưa download
import numpy as np
np.random.seed(42)
# === SAMPLE: E-commerce Dataset (10,000 orders) ===
n = 10000
orders = pd.DataFrame({
'order_id': [f'ORD_{i:05d}' for i in range(n)],
'customer_id': [f'CUST_{np.random.randint(1, 5000):05d}' for _ in range(n)],
'order_date': pd.date_range('2024-01-01', periods=n, freq='52min'),
'product_category': np.random.choice(
['Electronics', 'Fashion', 'Home', 'Beauty', 'Sports',
'Books', 'Food', 'Toys', 'Health', 'Auto'],
n, p=[0.15, 0.20, 0.12, 0.10, 0.08, 0.10, 0.08, 0.05, 0.07, 0.05]
),
'order_value': np.round(np.random.lognormal(4.5, 0.8, n), 2),
'quantity': np.random.choice([1, 2, 3, 4, 5], n, p=[0.45, 0.25, 0.15, 0.10, 0.05]),
'payment_method': np.random.choice(
['Credit Card', 'Debit Card', 'Bank Transfer', 'E-wallet', 'COD'],
n, p=[0.35, 0.20, 0.15, 0.15, 0.15]
),
'delivery_days': np.random.choice(range(1, 31), n),
'review_score': np.random.choice([1, 2, 3, 4, 5], n, p=[0.05, 0.08, 0.15, 0.32, 0.40]),
'city': np.random.choice(
['Ho Chi Minh', 'Ha Noi', 'Da Nang', 'Can Tho', 'Hai Phong',
'Nha Trang', 'Hue', 'Bien Hoa', 'Vung Tau', 'Buon Ma Thuot'],
n, p=[0.30, 0.25, 0.10, 0.08, 0.07, 0.05, 0.05, 0.04, 0.03, 0.03]
),
'customer_segment': np.random.choice(
['New', 'Regular', 'VIP', 'At-Risk', 'Churned'],
n, p=[0.25, 0.35, 0.15, 0.15, 0.10]
)
})
# Inject realistic missing values
mask_delivery = np.random.random(n) < 0.03 # 3% missing delivery
orders.loc[mask_delivery, 'delivery_days'] = np.nan
mask_review = np.random.random(n) < 0.05 # 5% missing reviews
orders.loc[mask_review, 'review_score'] = np.nan
print(f"✅ Dataset created: {orders.shape[0]:,} orders × {orders.shape[1]} columns")
print(f"📅 Date range: {orders['order_date'].min()} → {orders['order_date'].max()}")
orders.head()Phần 2: Data Audit — Profile & Evaluate (25 phút)
Bước 2.1: Data Profiling
python
# ============================================
# PHẦN 2: DATA AUDIT — PROFILE DATASET
# ============================================
print("=" * 65)
print("📊 DATA AUDIT REPORT")
print("=" * 65)
# --- 1. SHAPE ---
print(f"\n{'─' * 65}")
print(f"📐 1. SHAPE")
print(f"{'─' * 65}")
print(f" Rows: {orders.shape[0]:,}")
print(f" Columns: {orders.shape[1]}")
print(f" Pass ≥ 1,000 rows: {'✅ YES' if orders.shape[0] >= 1000 else '❌ NO'}")
# --- 2. DATA TYPES ---
print(f"\n{'─' * 65}")
print(f"📋 2. DATA TYPES")
print(f"{'─' * 65}")
for col in orders.columns:
print(f" {col:25s} → {str(orders[col].dtype):15s} | Unique: {orders[col].nunique():,}")
# --- 3. MISSING VALUES ---
print(f"\n{'─' * 65}")
print(f"❓ 3. MISSING VALUES")
print(f"{'─' * 65}")
missing = orders.isnull().sum()
missing_pct = (missing / len(orders) * 100).round(2)
total_missing = missing.sum()
total_cells = orders.shape[0] * orders.shape[1]
overall_missing_pct = (total_missing / total_cells * 100)
for col in orders.columns:
status = "✅" if missing[col] == 0 else ("⚠️" if missing_pct[col] < 10 else "❌")
print(f" {status} {col:25s} → {missing[col]:5,} missing ({missing_pct[col]:5.2f}%)")
print(f"\n Overall: {total_missing:,} / {total_cells:,} ({overall_missing_pct:.2f}%)")
print(f" Pass ≤ 30%: {'✅ YES' if overall_missing_pct <= 30 else '❌ NO'}")
# --- 4. NUMERIC SUMMARY ---
print(f"\n{'─' * 65}")
print(f"📈 4. NUMERIC SUMMARY")
print(f"{'─' * 65}")
numeric_cols = orders.select_dtypes(include=[np.number]).columns
for col in numeric_cols:
stats = orders[col].describe()
print(f" {col}:")
print(f" Mean: {stats['mean']:.2f} | Std: {stats['std']:.2f}")
print(f" Min: {stats['min']:.2f} | Max: {stats['max']:.2f}")
print(f" Q1: {stats['25%']:.2f} | Median: {stats['50%']:.2f} | Q3: {stats['75%']:.2f}")Bước 2.2: Categorical Summary
python
# --- 5. CATEGORICAL SUMMARY ---
print(f"\n{'─' * 65}")
print(f"🏷️ 5. CATEGORICAL SUMMARY")
print(f"{'─' * 65}")
cat_cols = orders.select_dtypes(include='object').columns
for col in cat_cols:
print(f"\n 📂 {col} ({orders[col].nunique()} unique):")
top5 = orders[col].value_counts().head(5)
for val, count in top5.items():
pct = count / len(orders) * 100
bar = "█" * int(pct / 2)
print(f" {val:20s} → {count:5,} ({pct:5.1f}%) {bar}")Bước 2.3: Duplicate & Date Check
python
# --- 6. DUPLICATES ---
print(f"\n{'─' * 65}")
print(f"🔄 6. DUPLICATES")
print(f"{'─' * 65}")
dups = orders.duplicated().sum()
print(f" Exact duplicates: {dups} ({dups/len(orders)*100:.2f}%)")
print(f" Pass: {'✅ No duplicates' if dups == 0 else f'⚠️ {dups} duplicates found'}")
# --- 7. DATE RANGE ---
print(f"\n{'─' * 65}")
print(f"📅 7. DATE RANGE")
print(f"{'─' * 65}")
if 'order_date' in orders.columns:
orders['order_date'] = pd.to_datetime(orders['order_date'])
date_range = (orders['order_date'].max() - orders['order_date'].min()).days
print(f" Start: {orders['order_date'].min()}")
print(f" End: {orders['order_date'].max()}")
print(f" Span: {date_range} days ({date_range/30:.1f} months)")
print(f" Has temporal data: ✅ YES")Bước 2.4: Dataset Viability Score
python
# ============================================
# DATASET VIABILITY SCORE (10-point checklist)
# ============================================
print(f"\n{'=' * 65}")
print(f"📋 DATASET VIABILITY CHECKLIST")
print(f"{'=' * 65}")
# Evaluate each criterion
checks = {
'1. Size ≥ 1,000 rows': orders.shape[0] >= 1000,
'2. ≥ 8 columns (mix types)': orders.shape[1] >= 8,
'3. Has date/time column': 'order_date' in orders.columns,
'4. Missing ≤ 30% overall': overall_missing_pct <= 30,
'5. Has documentation': True, # Manual check
'6. Business context clear': True, # Manual check
'7. Has target variable': True, # Manual check (review_score, segment)
'8. License OK': True, # Manual check
'9. Data ≤ 5 years old': True, # Manual check
'10. Not overused': True, # Manual check
}
score = sum(checks.values())
for criterion, passed in checks.items():
status = "✅" if passed else "❌"
print(f" {status} {criterion}")
print(f"\n{'─' * 65}")
print(f" 📊 SCORE: {score}/10")
print(f" {'✅ PASS — Dataset suitable for capstone!' if score >= 7 else '❌ FAIL — Find another dataset'}")
print(f"{'=' * 65}")Phần 3: Viết Project Brief (15 phút)
Bước 3.1: Create Project Brief
python
# ============================================
# PHẦN 3: PROJECT BRIEF
# ============================================
project_brief = """
╔═══════════════════════════════════════════════════════════════╗
║ 📋 PROJECT BRIEF ║
╠═══════════════════════════════════════════════════════════════╣
║ Title: E-commerce Customer Segmentation & Revenue Insights ║
║ Author: [YOUR NAME] ║
║ Date: [TODAY'S DATE] ║
╠═══════════════════════════════════════════════════════════════╣
║ 🎯 BUSINESS PROBLEM: ║
║ An e-commerce marketplace needs to understand customer ║
║ behavior to optimize marketing spend, improve delivery ║
║ experience, and identify high-value customer segments. ║
║ ║
║ 📊 DATASET: ║
║ - Source: Kaggle / Generated sample ║
║ - Size: {rows:,} rows × {cols} columns ║
║ - Period: {start} → {end} ║
║ - Key variables: order_value, product_category, ║
║ delivery_days, review_score, customer_segment, city ║
║ ║
║ ❓ BUSINESS QUESTIONS: ║
║ 1. Which customer segments have highest LTV? ║
║ → Marketing strategy per segment ║
║ 2. How does delivery time affect review ratings? ║
║ → Delivery threshold for negative experience ║
║ 3. What are revenue trends & seasonal patterns? ║
║ → Inventory and marketing planning ║
║ 4. Which product categories are growing fastest? ║
║ → Product mix optimization ║
║ 5. What factors predict customer churn? ║
║ → Retention strategy for at-risk customers ║
║ ║
║ 🔧 METHODS & TOOLS: ║
║ - Analysis: EDA, RFM Segmentation, Correlation, ║
║ Trend Analysis, optional ML (Logistic Regression) ║
║ - Tools: Python (pandas, matplotlib, seaborn), SQL, ║
║ Tableau / Power BI ║
║ - Libraries: pandas, numpy, matplotlib, seaborn, ║
║ scikit-learn (optional) ║
║ ║
║ 📦 DELIVERABLES: ║
║ - [x] Jupyter Notebooks (analysis + code) ║
║ - [ ] Dashboard (Tableau / Power BI) ║
║ - [ ] Presentation (10-15 slides) ║
║ - [x] GitHub README (project documentation) ║
║ ║
║ 📅 TIMELINE: ║
║ - Week 1: Data cleaning + EDA + Segmentation analysis ║
║ - Week 2: Dashboard + Presentation + Polish repo ║
╚═══════════════════════════════════════════════════════════════╝
""".format(
rows=orders.shape[0],
cols=orders.shape[1],
start=orders['order_date'].min().strftime('%Y-%m-%d'),
end=orders['order_date'].max().strftime('%Y-%m-%d')
)
print(project_brief)Bước 3.2: Questions → Methods Mapping
python
# ============================================
# QUESTIONS → METHODS MAPPING
# ============================================
mapping = pd.DataFrame({
'Question': [
'Q1: Customer segments by LTV?',
'Q2: Delivery → Rating impact?',
'Q3: Revenue trends & seasonality?',
'Q4: Fastest growing categories?',
'Q5: Churn prediction factors?'
],
'Method': [
'RFM Analysis + Clustering',
'Correlation + Threshold Analysis',
'Time Series Decomposition',
'Growth Rate + Trend Analysis',
'Logistic Regression (optional)'
],
'Tool': [
'Python (pandas, sklearn)',
'Python (seaborn, scipy)',
'Python (matplotlib, statsmodels)',
'Python (pandas) + Tableau',
'Python (sklearn)'
],
'Output': [
'Segment profiles + marketing recs',
'Delivery threshold + improvement recs',
'Forecast chart + seasonal calendar',
'Category growth dashboard',
'Risk factors + retention strategy'
],
'Priority': ['P0', 'P0', 'P1', 'P1', 'P2']
})
print("📋 QUESTIONS → METHODS → TOOLS → OUTPUT MAPPING")
print("=" * 80)
print(mapping.to_string(index=False))Phần 4: Setup Git Repo + README (25 phút)
Bước 4.1: Git Init & Folder Structure
Mở terminal và chạy từng lệnh:
bash
# ============================================
# PHẦN 4: GIT SETUP
# ============================================
# 1. Cấu hình Git (lần đầu tiên)
git config --global user.name "Tên Của Bạn"
git config --global user.email "email@example.com"
# 2. Tạo project folder
mkdir capstone-ecommerce-analysis
cd capstone-ecommerce-analysis
# 3. Init Git repository
git init
# 4. Tạo folder structure (TDSP-inspired)
mkdir -p data/raw
mkdir -p data/processed
mkdir -p notebooks
mkdir -p dashboards
mkdir -p reports/figures
mkdir -p src
# 5. Verify structure
# Windows:
dir /s /b
# Mac/Linux:
# find . -type dBước 4.2: Create .gitignore
bash
# ============================================
# CREATE .gitignore
# ============================================
# Tạo file .gitignore (Windows PowerShell)
@"
# === Data files (too large for Git) ===
*.csv
*.xlsx
*.parquet
*.json
data/raw/
# === Jupyter ===
.ipynb_checkpoints/
*.ipynb_metadata/
# === Python ===
__pycache__/
*.pyc
*.pyo
.env
venv/
.venv/
*.egg-info/
# === OS ===
.DS_Store
Thumbs.db
desktop.ini
~$*
# === IDE ===
.vscode/
.idea/
*.swp
*.swo
# === Tableau ===
*.twbx
*.hyper
"@ | Out-File -Encoding utf8 .gitignore
# Hoặc trên Mac/Linux:
# cat > .gitignore << 'EOF'
# [nội dung tương tự]
# EOFBước 4.3: Create README.md
bash
# ============================================
# CREATE README.md
# ============================================
# Tạo README.md (dùng text editor hoặc VS Code)
# code README.mdNội dung README.md:
markdown
# 📊 E-commerce Customer Segmentation & Revenue Insights
> Phân tích 10,000+ orders từ e-commerce marketplace — segment customers
> theo RFM, analyze delivery impact on ratings, identify growth categories.
<!--  -->
## 🎯 Business Problem
An e-commerce marketplace needs to understand customer behavior to:
- **Optimize marketing spend** by targeting the right customer segments
- **Improve delivery experience** by identifying pain point thresholds
- **Forecast revenue** for better inventory and budget planning
## 📊 Dataset
| Attribute | Detail |
|-----------|--------|
| **Source** | Kaggle / Generated sample |
| **Size** | 10,000 rows × 11 columns |
| **Period** | Jan 2024 — Dec 2024 |
| **Key Variables** | order_value, product_category, delivery_days, review_score, customer_segment |
## ❓ Key Questions
1. Which customer segments (RFM) have the highest LTV, and what marketing strategy suits each?
2. How does delivery time affect review ratings? What's the negative experience threshold?
3. What are the 12-month revenue trends and seasonal patterns?
4. Which product categories are growing fastest?
5. What factors predict customer churn?
## 🔧 Tools & Methods
- **Python**: pandas, numpy, matplotlib, seaborn, scikit-learn
- **SQL**: Data extraction and aggregation
- **BI**: Tableau / Power BI interactive dashboard
- **Methods**: EDA, RFM Segmentation, Correlation Analysis, Trend Analysis
## 📈 Key Findings
🔄 *In progress — will be updated after analysis*
## 💡 Recommendations
🔄 *In progress — will be updated after analysis*
## 📁 Project Structurecapstone-ecommerce-analysis/ ├── README.md ├── .gitignore ├── requirements.txt ├── data/ │ ├── raw/ # Original data (not tracked) │ └── processed/ # Cleaned data ├── notebooks/ │ ├── 01_data_audit.ipynb │ ├── 02_data_cleaning.ipynb │ ├── 03_eda.ipynb │ ├── 04_segmentation.ipynb │ └── 05_analysis.ipynb ├── dashboards/ │ └── dashboard_screenshot.png ├── reports/ │ ├── figures/ │ └── presentation.pdf └── src/ └── utils.py
## 🚀 How to Run
1. Clone: `git clone https://github.com/[username]/capstone-ecommerce-analysis.git`
2. Install: `pip install -r requirements.txt`
3. Run notebooks in order: 01 → 02 → 03 → 04 → 05
## 👤 Author
**[Your Name]** — Aspiring Data Analyst
- 🔗 [LinkedIn](https://linkedin.com/in/your-profile)
- 📧 your.email@example.comBước 4.4: Create requirements.txt
bash
# ============================================
# CREATE requirements.txt
# ============================================
# Tạo requirements.txt
@"
pandas==2.1.4
numpy==1.26.2
matplotlib==3.8.2
seaborn==0.13.0
scikit-learn==1.3.2
jupyter==1.0.0
openpyxl==3.1.2
"@ | Out-File -Encoding utf8 requirements.txt
# Hoặc nếu đã install packages:
# pip freeze > requirements.txtBước 4.5: First Commit + Push
bash
# ============================================
# FIRST COMMIT + PUSH TO GITHUB
# ============================================
# 1. Check status
git status
# 2. Add all files
git add .
# 3. First commit
git commit -m "feat: initial project setup — folder structure, README, .gitignore, requirements"
# 4. Verify commit
git log --oneline
# 5. Create GitHub repo (qua browser: github.com/new)
# Tên: capstone-ecommerce-analysis
# Visibility: Public
# Do NOT add README/gitignore (đã tạo locally)
# 6. Connect + Push
git remote add origin https://github.com/[username]/capstone-ecommerce-analysis.git
git branch -M main
git push -u origin main
# 7. Verify: mở browser → github.com/[username]/capstone-ecommerce-analysisBước 4.6: Save Data Audit to Repo
python
# ============================================
# SAVE DATA AUDIT REPORT
# ============================================
# Save dataset (processed, for reference)
orders.to_csv('data/processed/orders_sample.csv', index=False)
print("✅ Dataset saved: data/processed/orders_sample.csv")
# Save data audit report as markdown
audit_report = f"""# 📊 Data Audit Report
## Dataset Overview
- **Source:** Kaggle / Generated sample
- **Size:** {orders.shape[0]:,} rows × {orders.shape[1]} columns
- **Date range:** {orders['order_date'].min().strftime('%Y-%m-%d')} → {orders['order_date'].max().strftime('%Y-%m-%d')}
## Data Types
| Column | Type | Unique | Missing |
|--------|------|--------|---------|
"""
for col in orders.columns:
audit_report += f"| {col} | {orders[col].dtype} | {orders[col].nunique():,} | {orders[col].isnull().sum()} ({orders[col].isnull().mean()*100:.1f}%) |\n"
audit_report += f"""
## Viability Score
- **Score:** {score}/10
- **Result:** {'✅ PASS' if score >= 7 else '❌ FAIL'}
## Notes
- Delivery days: {orders['delivery_days'].isnull().sum()} missing values ({orders['delivery_days'].isnull().mean()*100:.1f}%) → impute with median
- Review score: {orders['review_score'].isnull().sum()} missing values ({orders['review_score'].isnull().mean()*100:.1f}%) → drop or impute
- Date range covers ~{(orders['order_date'].max() - orders['order_date'].min()).days} days
"""
with open('data/data_dictionary.md', 'w', encoding='utf-8') as f:
f.write(audit_report)
print("✅ Data audit report saved: data/data_dictionary.md")bash
# Commit data audit
git add data/data_dictionary.md data/processed/orders_sample.csv
git commit -m "data: add sample dataset and data audit report"
git pushPhần 5: Verification — Kiểm tra Final (5 phút)
Bước 5.1: Final Checklist
python
# ============================================
# PHẦN 5: FINAL VERIFICATION
# ============================================
import os
# Check all required files exist
required_files = [
'README.md',
'.gitignore',
'requirements.txt',
'data/data_dictionary.md',
'data/processed/orders_sample.csv',
]
required_dirs = [
'data/raw',
'data/processed',
'notebooks',
'dashboards',
'reports/figures',
'src',
]
print("=" * 60)
print("📋 FINAL VERIFICATION CHECKLIST")
print("=" * 60)
print("\n📄 Required Files:")
for f in required_files:
exists = os.path.exists(f)
print(f" {'✅' if exists else '❌'} {f}")
print("\n📁 Required Directories:")
for d in required_dirs:
exists = os.path.isdir(d)
print(f" {'✅' if exists else '❌'} {d}")
# Check Git status
print("\n🔧 Git Status:")
print(" Run: git status")
print(" Run: git log --oneline")
print(" Run: git remote -v")
print("\n" + "=" * 60)
print("🎯 DAY 1 COMPLETE CHECKLIST:")
print("=" * 60)
checklist = [
"☐ Đề tài chọn xong (DRBST ≥ 4/5)",
"☐ Dataset downloaded + audited (≥ 7/10)",
"☐ Project brief written (5 questions + methods)",
"☐ Git repo initialized + pushed to GitHub",
"☐ README.md v1 live on GitHub",
"☐ .gitignore + requirements.txt committed",
"☐ Data audit report saved",
"☐ Folder structure TDSP-style",
]
for item in checklist:
print(f" {item}")
print(f"\n✅ Ready for Buổi 19: Data Cleaning + EDA + Analysis!")📊 Output mẫu — Bạn cần có
| Deliverable | File/Location | Status |
|---|---|---|
| Git repo | github.com/[user]/capstone-ecommerce-analysis | ☐ Live |
| README.md | Root of repo | ☐ v1 committed |
| .gitignore | Root of repo | ☐ Committed |
| requirements.txt | Root of repo | ☐ Committed |
| Data audit report | data/data_dictionary.md | ☐ Committed |
| Dataset | data/processed/orders_sample.csv | ☐ Committed |
| Project brief | Notebook or document | ☐ Written |
| Folder structure | See TDSP layout | ☐ Created |
🎯 Tiếp theo — Buổi 19
Với repo setup và data audited, buổi tiếp theo bạn sẽ:
- Data cleaning: Handle missing, duplicates, types
- EDA: Distributions, correlations, key patterns
- Analysis: Answer business questions 1-5
- Dashboard: Build interactive BI dashboard
Day 1 (Today): ✅ Kickoff — dataset + repo + brief
Day 2 (Next): → Data cleaning + wrangling
Day 3-4: → EDA + visualizations
Day 5-7: → Deep analysis + answer questions
Day 8-9: → Dashboard
Day 10-14: → Presentation + polish💡 Commit mỗi ngày!
Mỗi session làm việc, commit ít nhất 1 lần. Green squares trên GitHub = evidence bạn đang active:
bash
git add .
git commit -m "feat: [mô tả ngắn gọn]"
git push