Python Database Data Analysis for making indicator
Budget: ₹12,500 – ₹37,500 INR
I’m looking for a Python-savvy analyst to turn the numerical records stored in my databases into clear, actionable insights. The job centres on data analysis (not web development or automation scripting), so I need someone comfortable pulling figures directly from relational tables, cleaning them with pandas, and exploring them in depth.
Here’s what I expect:
• Connect to the existing databases (PostgreSQL and MySQL are the main ones) using Python, SQLAlchemy or your preferred connector.
• Extract and clean the numerical fields, handling missing values and obvious outliers.
• Produce an organised, well-commented analysis script or Jupyter notebook that walks through descriptive statistics, key correlations, and at least one statistical test or model relevant to the dataset.
• Summarise findings in a concise report or notebook section with easy-to-read charts (matplotlib or seaborn is fine).
• Deliver all code, a short “how to run” README, and any requirements.txt.
I already have database credentials and can provide table schemas on kickoff. If you have a solid background in Python data analysis and SQL, I’d love to see a quick sample of similar work when you reply.
Here’s what I expect:
• Connect to the existing databases (PostgreSQL and MySQL are the main ones) using Python, SQLAlchemy or your preferred connector.
• Extract and clean the numerical fields, handling missing values and obvious outliers.
• Produce an organised, well-commented analysis script or Jupyter notebook that walks through descriptive statistics, key correlations, and at least one statistical test or model relevant to the dataset.
• Summarise findings in a concise report or notebook section with easy-to-read charts (matplotlib or seaborn is fine).
• Deliver all code, a short “how to run” README, and any requirements.txt.
I already have database credentials and can provide table schemas on kickoff. If you have a solid background in Python data analysis and SQL, I’d love to see a quick sample of similar work when you reply.