Data preprocessing¶
Convert and replace the data in a table. These methods are available on pykx.Table and pykx.KeyedTable objects.
astype ¶
astype(dtype, copy=True, errors='raise')
Cast columns of the table to a different data type.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
dtype |
type or dict
|
A |
required |
copy |
bool
|
Only |
True
|
errors |
str
|
|
'raise'
|
Returns:
| Type | Description |
|---|---|
Table
|
The table with the columns cast. |
Examples:
>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.astype(kx.FloatVector)
pykx.Table(pykx.q('
a b
---
1 4
2 5
3 6
'))
select_dtypes ¶
select_dtypes(include=None, exclude=None)
Select the columns of the table by their data types.
At least one of include and exclude is required and they cannot overlap. If both are
given exclude is ignored. For a KeyedTable the key columns are always kept and only the
value columns are filtered. Use kx.CharAtom for a column of single characters, as
kx.CharVector raises an error.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
include |
type, str or list
|
A data type or string, or a list of them, of the columns to keep. |
None
|
exclude |
type, str or list
|
A data type or string, or a list of them, of the columns to remove. |
None
|
Returns:
| Type | Description |
|---|---|
Table
|
The table of the selected columns, or |
Examples:
>>> tab = kx.q('([] a:1 2 3; b:1.5 2.5 3.5)')
>>> tab.select_dtypes(include=kx.FloatVector)
pykx.Table(pykx.q('
b
---
1.5
2.5
3.5
'))
replace ¶
replace(to_replace, value)
Replace all occurrences of a value in the table.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
to_replace |
Any
|
The value to replace. |
required |
value |
Any
|
The value to replace it with. |
required |
Returns:
| Type | Description |
|---|---|
Table
|
The table with the values replaced. |
Examples:
>>> tab = kx.q('([] a:1 2 3; b:4 5 6)')
>>> tab.replace(5, 50)
pykx.Table(pykx.q('
a b
----
1 4
2 50
3 6
'))