Groupby.col

Obtain the column(s) per groups
danfo.Groupby.col(col_names)    [source]
Parameters
Type
Description
Default
col_names
Array
List of column
​
Returns: Groupby Data structure
Note:    This is similar to pandas df.groupby(["column"])["col_names"]
Examples
Obtain a column
Node
Node
const dfd = require("danfojs-node")
​
let data ={'A': ['foo', 'bar', 'foo', 'bar',
                'foo', 'bar', 'foo', 'foo'],
           'B': ['one', 'one', 'two', 'three',
                'two', 'two', 'one', 'three'],
           'C': [1,3,2,4,5,2,6,7],
           'D': [3,2,4,1,5,6,7,8]
        }
​
let df = new dfd.DataFrame(data)
​
​
let grp = df.groupby(["A"])
grp.col(["C"])
​
//for more coumns
grp.col(["C","D"])
Apparently the output are not that useful unless you perform some operations like max(), count() and the likes. 
//it returns the groupby data structure
GroupBy {
  key_col: [ 'A' ],
  col_dict: {
    foo: [ [Array], [Array], [Array], [Array], [Array] ],
    bar: [ [Array], [Array], [Array] ]
  },
  data: [
    [ 'foo', 'one', 1, 3 ],
    [ 'bar', 'one', 3, 2 ],
    [ 'foo', 'two', 2, 4 ],
    [ 'bar', 'three', 4, 1 ],
    [ 'foo', 'two', 5, 5 ],
    [ 'bar', 'two', 2, 6 ],
    [ 'foo', 'one', 6, 7 ],
    [ 'foo', 'three', 7, 8 ]
  ],
  column_name: [ 'A', 'B', 'C', 'D' ],
  data_tensors: {
    foo: DataFrame {
      kwargs: [Object],
      series: false,
      data: [Array],
      row_data_tensor: [Tensor],
      index_arr: [Array],
      columns: [Array],
      col_data: [Array],
      col_data_tensor: [Tensor],
      col_types: [Array],
      A: [Getter/Setter],
      B: [Getter/Setter],
      C: [Getter/Setter],
      D: [Getter/Setter]
    },
    bar: DataFrame {
      kwargs: [Object],
      series: false,
      data: [Array],
      row_data_tensor: [Tensor],
      index_arr: [Array],
      columns: [Array],
      col_data: [Array],
      col_data_tensor: [Tensor],
      col_types: [Array],
      A: [Getter/Setter],
      B: [Getter/Setter],
      C: [Getter/Setter],
      D: [Getter/Setter]
    }
  },
  group_col_name: [ 'C' ],
  group_col: { foo: [ [Series] ], bar: [ [Series] ] }
}
​
Groupby.get_groups
Groupby.max
Last updated 5 months ago
Parameters	Type	Description	Default
col_names	Array	List of column