Skip to content

Formateurs & Code

CSV → JSON Dataset

CSV + inferred column metadata for ML pipelines.

Runs in your browser
JSON dataset with metadata
{
  "metadata": {
    "rowCount": 5,
    "columnCount": 6,
    "columns": [
      {
        "name": "id",
        "type": "integer",
        "nullCount": 0,
        "uniqueCount": 5
      },
      {
        "name": "name",
        "type": "string",
        "nullCount": 0,
        "uniqueCount": 5
      },
      {
        "name": "email",
        "type": "string",
        "nullCount": 0,
        "uniqueCount": 1
      },
      {
        "name": "age",
        "type": "integer",
        "nullCount": 0,
        "uniqueCount": 5
      },
      {
        "name": "active",
        "type": "boolean",
        "nullCount": 0,
        "uniqueCount": 2
      },
      {
        "name": "city",
        "type": "string",
        "nullCount": 0,
        "uniqueCount": 5
      }
    ]
  },
  "data": [
    {
      "id": 1,
      "name": "Ada",
      "email": "[email protected]",
      "age": 36,
      "active": true,
      "city": "London"
    },
    {
      "id": 2,
      "name": "Grace",
      "email": "[email protected]",
      "age": 85,
      "active": true,
      "city": "New York"
    },
    {
      "id": 3,
      "name": "Alan",
      "email": "[email protected]",
      "age": 41,
      "active": false,
      "city": "Cambridge"
    },
    {
      "id": 4,
      "name": "Linus",
      "email": "[email protected]",
      "age": 55,
      "active": true,
      "city": "Helsinki"
    },
    {
      "id": 5,
      "name": "Margaret",
      "email": "[email protected]",
      "age": 87,
      "active": true,
      "city": "Boston"
    }
  ]
}

Note de la rédaction

Understanding · Tabular rows into nested objects — for ML training and API responses.

Ce chapitre approfondi n'est actuellement disponible qu'en anglais. L'outil de conversion ci-dessus fonctionne dans votre langue ; le long article explicatif n'a pas encore été traduit.

Questions fréquentes

Quick answers.

›How does the tool handle data types?

The converter checks each value to see if it represents a number or a boolean. If a column consistently contains numeric patterns, it is output as a `number` type in the JSON rather than a string.

›Does my data leave my computer?

No. The parsing and transformation logic is executed locally within your browser session. No data is sent to our servers or stored in any database.

›What happens if my CSV has no header row?

The tool assumes the first row contains labels for the JSON keys. If headers are missing, the resulting objects will use the first row of data as keys, so it is best to provide a header-first file.

›Is there a limit on file size?

The limit depends on your browser's available memory. For datasets exceeding several hundred megabytes, we recommend using a dedicated local script to avoid browser tab crashes.

Les internautes recherchent aussi

Outils similaires

More in this room.

Tout voir dans Formateurs & Code