Added v1.0

A connector that can read and write training data to and from wrangles. Useful for updating wrangles training data without having to open Excel.

Tabset

Classify

Read

Recipe

read:
  - train.classify:
  		model_id: xxxxxxxx-xxxx-xxxx

Output

Example	Category	Notes
Apple	Fruit
Broccoli	Vegetable
Pear	Fruit
Aperagus	Vegetable

Parameters

Parameter	Required	Data Type	Notes
model_id	✓	str	Model id of wrangle to be read from
if		str	A condition that will determine whether the action runs or not as a whole.

Function

from wrangles.connectors import train
df = train.classify.read(model_id = 'xxxxxxxx-xxxx-xxxx')

Output

Example	Category	Notes
Apple	Fruit
Broccoli	Vegetable
Pear	Fruit
Asperagus	Vegetable

Parameters

Parameter	Required	Data Type	Notes
model_id	✓	str	Model id of wrangle to be read from
if		str	A condition that will determine whether the action runs or not as a whole.

Write

Recipe

Update the training data for an existing classify wrangle.

write:
  - train.classify:
  		columns:
        - Example
        - Category
        - Notes
  		model_id: xxxxxxxx-xxxx-xxxx

Parameters

Parameter	Required	Data Type	Notes
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Function

Create a new classify wrangle.

from wrangles.connectors import train
import pandas as pd

df = pd.DataFrame({
		'Example': ['Apple', 'Broccoli', 'Pear', 'Asperagus'],
    'Category': ['Fruit', 'Vegitable', 'Fruit', 'Vegitable'],
    'Notes': ['', '', '', '']
})

train.classify.write(df = df, name = 'Fruits and Veggies')

Parameters

Parameter	Required	Data Type	Notes
df	✓	Pandas DataFrame	Dataframe consisting of training data to be used
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Note both name and model_id cannot be used together. Name is used when creating models and model_id is used when updating an existing model.

Extract

Read

Recipe

read:
  - train.extract:
  		model_id: xxxxxxxx-xxxx-xxxx

Output

Entity to Find	Variation (Optional)	Notes
Ball Bearing	bearing, ball
Needle Bearing	needle, bearing
Roller Bearing	roller, bearing

Function

from wrangles.connectors import train
df = train.extract.read(model_id = 'xxxxxxxx-xxxx-xxxx')

Output

Entity to Find	Variation (Optional)	Notes
Ball Bearing	bearing, ball
Needle Bearing	needle, bearing
Roller Bearing	roller, bearing

Write

Recipe

Update the training data for an existing extract wrangle.

write:
  - train.extract:
  		columns:
        - Entity to Find
        - Variation (Optional)
        - Notes
  		model_id: xxxxxxxx-xxxx-xxxx

Parameters

Parameter	Required	Data Type	Notes
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Function

Create a new classify wrangle.

from wrangles.connectors import train
import pandas as pd

df = pd.DataFrame({
		'Entity to Find': ['Ball Bearing', 'Needle Bearing', 'Roller Bearing'],
    'Variation': ['bearing, ball', 'needle, bearing', 'roller, bearing'],
    'Notes': ['', '', '', '']
})

train.extract.write(df = df, name = 'Bearing Types')

Parameters

Parameter	Required	Data Type	Notes
df	✓	Pandas DataFrame	Dataframe consisting of training data to be used
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Note both name and model_id cannot be used together. Name is used when creating models and model_id is used when updating an existing model.

Lookup

Read

Recipe

read:
  - train.lookup:
  		model_id: xxxxxxxx-xxxx-xxxx

Output

Key	Value1	Value2
Pizza	Thin Crust	Deep Dish
Hamburger	Single	Double
Salad	Garden	Ceasar
Taco	Soft	Crispy

Function

from wrangles.connectors import train
df = train.lookup.read(model_id = 'xxxxxxxx-xxxx-xxxx')

Output

Key	Value1	Value2
Pizza	Thin Crust	Deep Dish
Hamburger	Single	Double
Salad	Garden	Ceasar
Taco	Soft	Crispy

Write

Recipe

Update the training data for an existing lookup wrangle.

write:
  - train.lookup:
  		model_id: xxxxxxxx-xxxx-xxxx

Parameters

Parameter	Required	Data Type	Notes
settings		dict	Specific settings to apply to the wrangle. Settings include variant which can be key or semantic. Settings must be used when creating a new lookup.
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
columns		str, list	The columns you wish to write to the Wrangle. Note: Columns must include one column named Key which will be used for the value to be looked up.
if		str	A condition that will determine whether the action runs or not as a whole.

Function

Create a new lookup wrangle.

from wrangles.connectors import train
import pandas as pd

df = pd.DataFrame({
		'Key': ['Pizza', 'Hamburger', 'Salad', 'Taco'],
    'Value1': ['Thin Crust', 'Single', 'Garden', 'Soft'],
    'Value2': ['Deep Dish', 'Double', 'Ceasar', 'Crispy']
})

train.lookup.write(df = df, name = 'Menu Updates', settings = {'variant': 'key'})

Parameters

Parameter	Required	Data Type	Notes
df	✓	Pandas DataFrame	Dataframe consisting of training data to be used
settings		dict	Specific settings to apply to the wrangle. Settings include variant which can be key or semantic. Settings must be used when creating a new lookup.
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Note both name and model_id cannot be used together. Name is used when creating models and model_id is used when updating an existing model.

Standardize

Read

Recipe

read:
  - train.standardize:
  		model_id: xxxxxxxx-xxxx-xxxx

Output

Find	Replace	Notes
Pizza	Hamburger
Beef	Chicken
Bagel	Donut
Taco	Burrito

Function

from wrangles.connectors import train
df = train.standardize.read(model_id = 'xxxxxxxx-xxxx-xxxx')

Output

Find	Replace	Notes
Pizza	Hamburger
Beef	Chicken
Bagel	Donut
Taco	Burrito

Write

Recipe

Update the training data for an existing standardize wrangle.

write:
  - train.standardize:
  		columns:
        - Find
        - Replace
        - Notes
  		model_id: xxxxxxxx-xxxx-xxxx

Parameters

Parameter	Required	Data Type	Notes
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Function

Create a new standardize wrangle.

from wrangles.connectors import train
import pandas as pd

df = pd.DataFrame({
		'Find': ['Pizza', 'Beef', 'Bagel', 'Taco'],
    'Replace': ['Hamburger', 'Chicken', 'Donut', 'Burrito'],
    'Notes': ['', '', '', '']
})

train.standardize.write(df = df, name = 'Menu Updates')

Parameters

Parameter	Required	Data Type	Notes
df	✓	Pandas DataFrame	Dataframe consisting of training data to be used
columns		list	List of columns to use for update/wrangle creation, defaults to none
model_id		str	Model id of wrangle to update. Only use for existing wrangles, defaults to none
name		str	Name of wrangle that is being created, defaults to none
if		str	A condition that will determine whether the action runs or not as a whole.

Note both name and model_id cannot be used together. Name is used when creating models and model_id is used when updating an existing model.