{
  "context": "This criteria evaluates how well the engineer uses the tabulate package's type inference capabilities to analyze column data types. The focus is on leveraging tabulate's internal type detection mechanisms rather than implementing custom type checking logic.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Uses tabulate internals",
      "description": "The solution imports tabulate and leverages its internal type detection mechanisms such as _column_type(), _type(), or similar internal functions rather than implementing custom type checking from scratch",
      "max_score": 30
    },
    {
      "name": "Type hierarchy",
      "description": "The solution correctly handles tabulate's type hierarchy where None < bool < int < float < str, promoting mixed types to the most general type (e.g., mixing int and float results in float type)",
      "max_score": 25
    },
    {
      "name": "Column transposition",
      "description": "The solution transposes row-based data into column-based data to analyze types per column, similar to how tabulate processes data internally with zip() or similar techniques",
      "max_score": 20
    },
    {
      "name": "None handling",
      "description": "The solution correctly ignores None values during type inference, ensuring they do not affect the determined column type as per tabulate's behavior",
      "max_score": 15
    },
    {
      "name": "Accurate type detection",
      "description": "The solution correctly identifies Python types (int, float, str, bool) and returns appropriate type labels matching tabulate's type detection logic",
      "max_score": 10
    }
  ]
}