Indexing and selection¶

Select rows, columns and values from a table, relabel it and set its index. These methods are available on pykx.Table and pykx.KeyedTable objects.

head ¶

head(n=5)

Return the first n rows, or all but the last -n rows if n is negative.

Examples:

>>> tab = kx.q('([] a:til 10)')
>>> tab.head(3)
pykx.Table(pykx.q('
a
-
0
1
2
'))

tail ¶

tail(n=5)

Return the last n rows, or all but the first -n rows if n is negative.

Examples:

>>> tab = kx.q('([] a:til 10)')
>>> tab.tail(3)
pykx.Table(pykx.q('
a
-
7
8
9
'))

pop ¶

pop(col_name)

Pop a column / columns from a table by name and return it.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.pop('b')
pykx.Table(pykx.q('
b
-
4
5
6
'))
>>> tab
pykx.Table(pykx.q('
a
-
1
2
3
'))

get ¶

get(key, default=None)

Get items from table based on key, if key is not found default is returned.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.get('a')
pykx.LongVector(pykx.q('1 2 3'))
>>> tab.get(['a', 'b'])
pykx.Table(pykx.q('
a b
---
1 4
2 5
3 6
'))

at `property` ¶

at()

Access or set a single value by row label and column name.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.at[1, 'b']
pykx.LongAtom(pykx.q('5'))
>>> tab.at[1, 'b'] = 50
>>> tab
pykx.Table(pykx.q('
a b
----
1 4
2 50
3 6
'))

loc `property` ¶

loc()

Access or set rows and columns by label or by a boolean mask.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.loc[tab['a'] > 1, 'b']
pykx.Table(pykx.q('
b
-
5
6
'))

iloc `property` ¶

iloc()

Access or set rows and columns by integer position.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.iloc[[0, 2], ['a']]
pykx.Table(pykx.q('
a
-
1
3
'))

drop ¶

drop(labels=None, axis=0, index=None, columns=None, level=None, inplace=False, errors='raise')

Remove rows or columns from a table.

Parameters:

Name	Type	Description	Default
`labels`	`str, int or list`	The row or column labels to remove, depending on `axis`.	`None`
`axis`	`int`	0 removes rows and 1 removes columns.	`0`
`index`	`list`	The row labels to remove, as an alternative to `labels` with `axis=0`.	`None`
`columns`	`str or list`	The column labels to remove, as an alternative to `labels` with `axis=1`.	`None`
`level`	`int or str`	The key level to remove rows from, for a keyed table with several key columns.	`None`
`inplace`	`bool`	Not implemented, must be `False`.	`False`
`errors`	`str`	`'raise'` raises an error if a label is not found and `'ignore'` skips it.	`'raise'`

Returns:

Type	Description
`Table`	The table with the rows or columns removed.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.drop(columns=['b'])
pykx.Table(pykx.q('
a
-
1
2
3
'))
>>> tab.drop([1])
pykx.Table(pykx.q('
a b
---
1 4
3 6
'))

drop_duplicates ¶

drop_duplicates(subset=None, keep='first', inplace=False, ignore_index=False)

Remove duplicate rows from a table.

Parameters:

Name	Type	Description	Default
`subset`	`None`	Not yet implemented, must be `None`.	`None`
`keep`	`str`	Only `'first'` is implemented.	`'first'`
`inplace`	`bool`	Not yet implemented, must be `False`.	`False`
`ignore_index`	`bool`	Not yet implemented, must be `False`.	`False`

Returns:

Type	Description
`Table`	The table with only the first of each set of duplicate rows. A `KeyedTable` is not supported.

Examples:

>>> tab = kx.q('([] a:1 1 2; b:3 3 4)')
>>> tab.drop_duplicates()
pykx.Table(pykx.q('
a b
---
1 3
2 4
'))

rename ¶

rename(
    mapper=None,
    index=None,
    columns=None,
    axis=0,
    copy=None,
    inplace=False,
    level=None,
    errors="ignore",
)

Rename the columns or the index of a table.

Parameters:

Name	Type	Description	Default
`mapper`	`dict`	A dictionary of names to new names, applied to the axis given by `axis`.	`None`
`index`	`dict`	A dictionary of index names to new names, for a keyed table with a single key column.	`None`
`columns`	`dict`	A dictionary of column names to new names.	`None`
`axis`	`int or str`	0 or `'index'` renames the index and 1 or `'columns'` renames the columns.	`0`
`copy`	`bool`	Not yet implemented, must be `None`.	`None`
`inplace`	`bool`	Not yet implemented, must be `False`.	`False`
`level`	`None`	Not yet implemented, must be `None`.	`None`
`errors`	`str`	Not yet implemented, must be `'ignore'`.	`'ignore'`

Returns:

Type	Description
`Table or KeyedTable`	The table with the names changed.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.rename(columns={'a': 'x'})
pykx.Table(pykx.q('
x b
---
1 4
2 5
3 6
'))

add_suffix ¶

add_suffix(suffix, axis=0)

Add a suffix to the name of each column.

Parameters:

Name	Type	Description	Default
`suffix`	`str`	The string added to the end of each column name.	required
`axis`	`int`	0 for a `Table` and 1 for the value columns of a `Table` or `KeyedTable`.	`0`

Returns:

Type	Description
`Table`	The table with the columns renamed.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.add_suffix('_x')
pykx.Table(pykx.q('
a_x b_x
-------
1   4
2   5
3   6
'))

add_prefix ¶

add_prefix(prefix, axis=0)

Add a prefix to the name of each column.

Parameters:

Name	Type	Description	Default
`prefix`	`str`	The string added to the start of each column name.	required
`axis`	`int`	0 for a `Table` and 1 for the value columns of a `Table` or `KeyedTable`.	`0`

Returns:

Type	Description
`Table`	The table with the columns renamed.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.add_prefix('x_')
pykx.Table(pykx.q('
x_a x_b
-------
1   4
2   5
3   6
'))

sample ¶

sample(
    n=None, frac=None, replace=False, weights=None, random_state=None, axis=None, ignore_index=False
)

Select random rows from the table.

Parameters:

Name	Type	Description	Default
`n`	`int`	The number of rows to return. This cannot be used with `frac` and defaults to 1.	`None`
`frac`	`float`	The fraction of the rows to return. This cannot be used with `n`.	`None`
`replace`	`bool`	Whether the same row can be selected more than once.	`False`
`weights`	`None`	Not yet implemented, must be `None`.	`None`
`random_state`	`None`	Not yet implemented, must be `None`.	`None`
`axis`	`int`	Not yet implemented, must be `None`.	`None`
`ignore_index`	`bool`	Not yet implemented, must be `False`.	`False`

Returns:

Type	Description
`Table`	A table of the selected rows.

Examples:

>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.sample(2)
pykx.Table(pykx.q('
a b
---
1 4
2 5
'))

set_index ¶

set_index(keys, drop=True, append=False, inplace=False, verify_integrity=False)

Set the index using existing columns.

Examples:

>>> tab = kx.q('([] k:`a`b`c; v:1 2 3)')
>>> tab.set_index('k')
pykx.KeyedTable(pykx.q('
k| v
-| -
a| 1
b| 2
c| 3
'))

index `property` ¶

index()

The index of the table, which is the key of a keyed table or the row numbers of a table.

Examples:

>>> tab = kx.q('([] k:`a`b`c; v:1 2 3)')
>>> tab.index
pykx.LongVector(pykx.q('0 1 2'))
>>> tab.set_index('k').index
pykx.Table(pykx.q('
k
-
a
b
c
'))

reset_index ¶

reset_index(
    levels=None,
    *,
    drop=False,
    inplace=False,
    col_level=0,
    col_fill="",
    allow_duplicates=False,
    names=None
)

Reset keys/index of a pykx.Table

Examples:

>>> tab = kx.q('1!([] k:`a`b`c; v:1 2 3)')
>>> tab.reset_index()
pykx.Table(pykx.q('
k v
---
a 1
b 2
c 3
'))

Indexing and selection¶

head ¶

tail ¶

pop ¶

get ¶

at property ¶

loc property ¶

iloc property ¶

drop ¶

drop_duplicates ¶

rename ¶

add_suffix ¶

add_prefix ¶

sample ¶

set_index ¶

index property ¶

reset_index ¶

at `property` ¶

loc `property` ¶

iloc `property` ¶

index `property` ¶