Extract a value from an association as a part:

Extract a value from an association by function application:

Extract a column from a Tabular object:

The Key wrapper is not necessary when the key is a string:

Lookups in associations interoperate with parts in lists and other expressions:

The explicit Key can be omitted when the keys in an association are strings:

Key is normally generated by functions such as Position:

Use Extract to extract the values:

For expressions at deeper levels, the positions can be combinations of integers and Key objects:

Use Key to disambiguate between an integer index and an integer key:

assoc[[Key[key]]] is equivalent to assoc[key], including integer keys:

Take a Tabular object with several columns:

Columns with non-string keys require the explicit use of Key for Part extraction:

String keys can be used with or without a Key wrapper:

Repeated keys are treated as one, even if they are wrapped with Key:

Take a Dataset object with several rows:

Use Part to extract a row with a string key:

Alternatively, use ds[key], which is equivalent to Query[key][ds]:

For non-string keys, extraction requires wrapping them with Key:

Otherwise, Query will interpret the key as a function to be used on the row:

Use Key in the Function body of Tabular transformations when the key is a general expression:

Key always takes one argument:

ExtendedKey can take more than one argument:

Both keys can then be used to address their values:

Take a Tabular object with a key column:

Key wrappers can be used to extract tabular rows and columns:

Row extraction can also be performed with RowKey: