Pulling Data from a Knowledge Model¶

In this tutorial, you will learn how to extract data using a knowledge model. More specifically, you will learn:

• How to connect to the EMS
• How to get and interact with a knowledge model and its objects
• How to create a PQL query using knowledge model attributes
• How to extract data from the data model

Prerequisites¶

To follow this tutorial, you should have PyCelonis installed and should know how to perform basic interactions with PyCelonis objects. If you don't know how to do this, please complete the Celonis Basics tutorial first. Further, it would be helpful to already have the previously mentioned assets inside your EMS. Please refer to the Studio - Introduction and Data Integration - Data Export tutorials for an overview of working with each asset type.

1. Import PyCelonis and connect to Celonis API¶

In [1]:

from pycelonis.pql import PQL, PQLColumn, PQLFilter
from pycelonis import get_celonis
celonis = get_celonis()

from pycelonis.pql import PQL, PQLColumn, PQLFilter from pycelonis import get_celonis celonis = get_celonis()

[2023-02-07 22:14:16,478] INFO: Initial connect successful! PyCelonis Version: 2.0.1

2. Get the knowledge model, record, and filter objects¶

In [2]:

space = celonis.studio.get_spaces().find('Space_Name_Here')
package = space.get_packages().find('Package_Name_Here')
knowledge_model = package.get_knowledge_model('Knowledge_Model_Name_Here')
content = knowledge_model.get_content()
data_model_id = content.data_model_id

record_name = 'Record_Name_Here'
records = content.records
record = records.find(search_term= record_name, search_attribute='display_name')

filter_name = 'Filter_Name_Here'
filters = content.filters
filter = filters.find(search_term=filter_name, search_attribute='display_name')

space = celonis.studio.get_spaces().find('Space_Name_Here') package = space.get_packages().find('Package_Name_Here') knowledge_model = package.get_knowledge_model('Knowledge_Model_Name_Here') content = knowledge_model.get_content() data_model_id = content.data_model_id record_name = 'Record_Name_Here' records = content.records record = records.find(search_term= record_name, search_attribute='display_name') filter_name = 'Filter_Name_Here' filters = content.filters filter = filters.find(search_term=filter_name, search_attribute='display_name')

3. Create a PQL query using knowledge model attributes and filter¶

In [3]:

query = PQL()

attribute_names = ['Attribute Name 1 Here', 'Attribute Name 2 Here', 'Attribute Name 3', 'Attribute Name 4']
attributes = record.attributes

for name in attribute_names:
    attribute = attributes.find(name,search_attribute='display_name')
    query += PQLColumn(name=attribute.display_name, query=attribute.pql)

query += PQLFilter(name=filter.display_name, query=filter.pql)

query

query = PQL() attribute_names = ['Attribute Name 1 Here', 'Attribute Name 2 Here', 'Attribute Name 3', 'Attribute Name 4'] attributes = record.attributes for name in attribute_names: attribute = attributes.find(name,search_attribute='display_name') query += PQLColumn(name=attribute.display_name, query=attribute.pql) query += PQLFilter(name=filter.display_name, query=filter.pql) query

Out[3]:

PQL(columns=[
	PQLColumn(name='Client ID', query='"BSEG"."MANDT"'),
	PQLColumn(name='Doc. Number', query='"BSEG"."BELNR"'),
	PQLColumn(name='Company Code', query='"BSEG"."BUKRS"'),
	PQLColumn(name='Vendor Number', query='"LFA1"."LIFNR"')
], filters=[
	PQLFilter(query='FILTER LEN("BKPF"."XBLNR") > 4;')
], order_by_columns=[], distinct=False, limit=None, offset=None)

4. Get the data pool and data model objects¶

Option #1: Use the data pool name¶

In [ ]:

data_pool = celonis.data_integration.get_data_pools().find('Data_Pool_Name_Here')
data_model = data_pool.get_data_model(data_model_id)

data_pool = celonis.data_integration.get_data_pools().find('Data_Pool_Name_Here') data_model = data_pool.get_data_model(data_model_id)

Option #2: Use the data pool id (for Optimization)¶

In [4]:

# Use the data pool id (found in the data pool URL) to optimize your script
data_pool = celonis.data_integration.get_data_pool('Data_Pool_ID_Here')
data_model = data_pool.get_data_model(data_model_id)

# Use the data pool id (found in the data pool URL) to optimize your script data_pool = celonis.data_integration.get_data_pool('Data_Pool_ID_Here') data_model = data_pool.get_data_model(data_model_id)

5. Query the data model¶

In [5]:

import pycelonis.pql as pql
from pycelonis.pql.saola_connector import KnowledgeModelSaolaConnector

df = pql.DataFrame.from_pql(query, saola_connector=KnowledgeModelSaolaConnector(data_model, knowledge_model))
df.head()

import pycelonis.pql as pql from pycelonis.pql.saola_connector import KnowledgeModelSaolaConnector df = pql.DataFrame.from_pql(query, saola_connector=KnowledgeModelSaolaConnector(data_model, knowledge_model)) df.head()

[2023-02-07 22:14:52,256] INFO: Successfully created data export with id 'bc4d526c-b0f0-4eac-89bd-3b26db5157d1'
[2023-02-07 22:14:52,258] INFO: Wait for execution of data export with id 'bc4d526c-b0f0-4eac-89bd-3b26db5157d1'

0it [00:00, ?it/s]

[2023-02-07 22:14:52,323] INFO: Export result chunks for data export with id 'bc4d526c-b0f0-4eac-89bd-3b26db5157d1'

Out[5]:

	Client ID	Doc. Number	Company Code	Vendor Number
0	800_	0000109290_	3000_	0000001005_
1	800_	0000111453_	3000_	COMUNASA_
2	800_	0000111529_	3000_	0000100058_
3	800_	0000111588_	3000_	0000200026_
4	800_	0000111600_	3000_	0000200026_
...	...	...	...	...
3767	800_	0000075924_	3000_	0000001075_
3768	800_	0000075930_	3000_	0000003511_
3769	800_	0000076079_	3000_	0000001200_
3770	800_	0000076297_	3000_	0000001000_
3771	800_	0000000797_	3000_	0000001000_

3772 rows × 4 columns

Conclusion¶

Congratulations! You have learned how to connect to the EMS, get and interact with a knowledge model and its objects, create a PQL query using knowledge model attributes, and extract the data.