pycrunch 0.5.8

使用pycrunch

要在项目中使用pycrunch，请运行

$ python setup.py develop

这将使此目录中的代码对其他项目可用。

入门

通过以下方式启动简单的站点会话：

>>> import pycrunch
>>> site = pycrunch.connect(api_key="DFIJFIJWIEJIJFKSJLKKDJKFJSLLSLSL", site_url="https://your-domain.crunch.io/api/")

或者，如果您没有API密钥

>>> site = pycrunch.connect("me@mycompany.com", "yourpassword", "https://your-domain.crunch.io/api/")

然后，您可以创建一个API密钥

>>> apk = site.apikeys.create({"body": {"name": "API Key"}})
>>> apk.refresh()
>>> site_via_api_key = pycrunch.connect(api_key=apk.body["key"], site_url="https://your-domain.crunch.io/api/")

或者，如果您有Crunch访问令牌

>>> import pycrunch
>>> site = pycrunch.connect_with_token("DFIJFIJWIEJIJFKSJLKKDJKFJSLLSLSL", "https://your-domain.crunch.io/api/")

然后，您可以浏览网站。使用 print 来美化打印 JSON 负载数据

>>> print(site)
pycrunch.shoji.Catalog(**{
    "element": "shoji:catalog",
    "self": "https://your-domain.crunch.io/api/",
    "description": "The API root.",
    "catalogs": {
        "datasets": "https://your-domain.crunch.io/api/datasets/",
        ...
    },
    "urls": {
        "logout_url": "https://your-domain.crunch.io/api/logout/",
        ...
    },
    "views": {
        "migration": "https://your-domain.crunch.io/api/migration/"
    }
})

负载的目录、视图、片段和 URL 集合中的 URI 可以自动访问

>>> print(site.datasets)
pycrunch.shoji.Catalog(**{
    "self": "https://your-domain.crunch.io/api/datasets/",
    "element": "shoji:catalog",
    "index": {
        "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/": {
            "owner_display_name": "me@mycompany.com",
            "description": "",
            "id": "dbf9fca7b727",
            "owner_id": "https://your-domain.crunch.io/api/users/253b68/",
            "archived": false,
            "name": "Hog futures tracking (May 2014)"
        },
    },
    ...
})

每个识别的 JSON 负载还自动提供对每个 JSON 对象成员的点属性访问

>>> print(site.datasets.index.values()[0])
pycrunch.shoji.Tuple(**{
    "owner_display_name": "me@mycompany.com",
    "description": "",
    "id": "dbf9fca7b727",
    "owner_id": "https://your-domain.crunch.io/api/users/253b68/",
    "archived": false,
    "name": "Hog futures tracking (May 2014)"
})

响应还可以具有附加的帮助程序，例如目录索引中每个元组的 entity 属性，该属性遵循到实体资源的链接

>>> print(site.datasets.index.values()[0].entity_url)
"https://your-domain.crunch.io/api/datasets/dbf9fca7b727/"

>>> print(site.datasets.index.values()[0].entity)
pycrunch.shoji.Entity(**{
    "self": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/",
    "element": "shoji:entity",
    "description": "Detail for a given dataset",
    "body": {
        "archived": false,
        "user_id": "253b68",
        "name": "Hog futures tracking (May 2014)"
        "weight": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/variables/36f5404/",
        "creation_time": "2014-03-06T18:23:26.780752+00:00",
        "description": ""
    },
    "catalogs": {
        "batches": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/batches/",
        "joins": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/joins/",
        "variables": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/variables/",
        "filters": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/filters/",
        ...
    },
    "views": {
        "cube": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/cube/",
        ...
    },
    "urls": {
        "revision_url": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/revision/",
        ...
    },
    "fragments": {
        "table": "https://your-domain.crunch.io/api/datasets/dbf9fca7b727/table/"
    }
})

通常您通过其 create 方法将新资源添加到目录中

>>> ds = site.datasets.create({"body": {
        'name': "My first dataset"
    }}, refresh=True)
>>> gender = ds.variables.create({"body": {
        'name': 'Gender',
        'alias': 'gender',
        'type': 'categorical',
        'categories': [
            {'id': -1, 'name': 'No Data', 'numeric_value': None, 'missing': True},
            {'id': 1, 'name': 'M', 'numeric_value': None, 'missing': False},
            {'id': 2, 'name': 'F', 'numeric_value': None, 'missing': False}
        ],
        'values': [1, 2, {"?": -1}, 2]
    }}, refresh=True)
>>> print(ds.table.data)
pycrunch.elements.JSONObject(**{
    "e7f361628": [
        1,
        2,
        {"?": -1},
        2
    ]
})

要访问数据集中数据的 Pandas Dataframe

>>> from pycrunch import pandaslib as crunchpandas
>>> df = crunchpandas.dataframe_from_dataset(site,'baadf00d000339d9faadg00beab11e')
>>> print(df)
< Draws a dataframe table >