How cmattosr/bootcampds_projeto03 works
Project type and purpose
• This is a data-science analysis and forecasting project (not a web app or multi-service system).
File structure
• Top-level files:
• Projeto_Modulo03.ipynb – the main analysis notebook
• README.md – project description and instructions
• Directories:
• dados/ – the raw dataset (20210701.zip)
• imagens/ – any generated plots or static images
• No separate frontend/backend/services—everything runs in one notebook kernel.
Extract insights from the README
• Data source: Brasil.io COVID-19 dataset (cases, deaths, epidemiological week)
• Recommended steps: data cleaning, exploratory analysis, hypothesis formation, time-series forecasting, and conclusion.
• Deliverables: a coherent narrative notebook with well-configured graphs and clear conclusions.
a. Main components:
1. Data Ingestion
2. Data Cleaning & Transformation
3. Exploratory Data Analysis (EDA)
4. Time-Series Modeling & Forecasting
5. Visualization & Reporting
6. Inputs/Outputs (raw data ZIP, final plots, forecast tables)
b. Relationships & Data Flow:
• Outputs flow from Reporting → imagens/ & notebook outputs
c. Architectural patterns / design principles:
• Linear pipeline with feedback loops (iterate EDA ↔ cleaning)
• Single-process, pipeline-style orchestration within Jupyter
• Modular stages for reusability and clarity
d. Key technologies and libraries:
• Python & Jupyter Notebook (kernel orchestrates all steps)
• pandas / numpy (data manipulation)
• matplotlib / seaborn / plotly (visualization)
• statsmodels / Prophet / scikit-learn (time-series forecasting)
• Brasil.io API or direct CSV ingestion (data source)
Elements
• Blue for data movement (Ingestion → Cleaning → …)
• Green for analysis/modeling stages
• Yellow for outputs (plots, forecasts)
Additional guidance
• Important decisions (e.g., choice of forecasting library).