MAnalyzer: The Complete Guide to Features & Use Cases

Getting Started with MAnalyzer: Setup to First Analysis

1. System requirements

• OS: Windows ⁄₁₁, macOS 12+, or Ubuntu 20.04+
• RAM: 8 GB minimum (16 GB recommended for large datasets)
• Disk: 2 GB free for app + space for datasets
• Dependencies: Python 3.10+ (if using CLI/API), Java Runtime (if using desktop build)

2. Installation (assumes default desktop + CLI options)

1. Download installer for your OS from the official MAnalyzer site.
2. Run installer and follow prompts (choose Typical install).
3. (Optional CLI) Open terminal and install with pip:

   Code
   Copy CodeCopied
   pip install manalyzer 

4. Verify installation:

   Code
   Copy CodeCopied
   manalyzer –version 

3. Initial configuration

• Launch the app and create a workspace (name + storage location).
• Set preferred data directory in Settings.
• Configure default file formats (CSV, Parquet) and memory limits.
• (Optional) Connect to external data sources: databases (Postgres, MySQL), S3, or Google Drive — enter credentials in the Connections panel and test.

4. Importing your first dataset

1. Click Import → choose file or connection.
2. For CSV: specify delimiter, encoding, and whether first row is header.
3. Preview and adjust column types (string, numeric, datetime) before confirming.
4. Save as a project table.

5. Quick data check

• Run basic profiling: MAnalyzer → Profile Table.
  • View row count, missing values, distribution histograms, and summary stats.
• Use built-in data quality checks: duplicate detection, missing-value percentage, and outlier flags.

6. First simple analysis (exploratory)

1. Create a new Analysis → choose your table.
2. Add visualizations:
   • Histogram for numeric distributions.
   • Bar chart for categorical frequencies.
   • Time series for datetime trends.
3. Apply filters (e.g., date range or category) and group-by operations.
4. Use the Summary panel to inspect mean, median, std, and top categories.

7. Running a basic model (optional)

• Select Modeling → choose algorithm (linear regression or decision tree).
• Specify target column and features (auto-select or manual).
• Train with split (⁄₃₀) and review metrics (RMSE, R², accuracy).
• Export model or deployment artifact.

8. Exporting results

• Export cleaned data or analysis as CSV, Parquet, or Excel.
• Export visuals as PNG/SVG and analysis report as PDF.
• Save notebook-style workflow for reproducibility.

9. Short troubleshooting

• Import fails: check encoding and delimiter, increase upload chunk size.
• Slow performance: reduce in-memory preview rows, increase memory limit, or use sampling.
• Connection errors: verify credentials, whitelist IP, or test with a sample query.

10. Next steps (recommended)

• Run a full data quality pipeline and save as scheduled job.
• Try automating repetitive transforms with reusable macros.
• Explore advanced modeling (feature engineering, cross-validation).