RANDOM_FOREST

Calculate Forecast Values using Random Forest model
Special Method 004 4 params

Dataset dependencies extracted from SQL body

Reads from
None detected
Writes to
→ 004 000001 Historical opportunities and Pipeline param:P01

Used by 2 operations

DTPActivityOperationParameter valuesOrder
000001
Calculations & Data Validation
 040 020
Predicting WINLOS
P02 → lstyr_activities_cat,ly_purchase_nr_cat,inactivity_cat,ly_purchase_value_cat,weighted_value_cat,aging_cat,tot_activities_cat,lstmnt_activities_cat,lstqrt_activities_cat,product,type,prospect_size,winlos
P03 → winlos
P01 → 000001 Historical opportunities and Pipeline Training Data
P04 → VARCHAR(1)
1
000001
Calculations & Data Validation
 040 030
Predicting AGING
P02 → lstyr_activities_cat,ly_purchase_nr_cat,inactivity_cat,ly_purchase_value_cat,weighted_value_cat,aging_cat,tot_activities_cat,lstmnt_activities_cat,lstqrt_activities_cat,product,type,prospect_size,winlos
P03 → aging_cat
P04 → INTEGER
P01 → 000001 Historical opportunities and Pipeline Predictive Data
2

Parameters 4

CodeDescriptionType
P03 Field to predict S
P02 Fields List for Random Forest Model S
P01 Training Dataset T
P04 Type of Field to Predict S

PostgreSQL script PostgreSQL source
Sources: ows_metodo_speciale

Script anatomy — approximate, parsed from the SQL text
Sources R_RandomForestresult Outputs id · value
11 lines 
CREATE TEMP TABLE result (id VARCHAR(36), value &P04&);

INSERT INTO result(id, value)
select
OID AS id ,
CAST(predValid AS &P04& ) AS value
from R_RandomForest('&P01&', '&P02&', '&P03&', '#TARGET_PARTITION#') as t(OID text, predValid text);

EXECUTE 'UPDATE #TARGET_PARTITION# SET  &P03& = result.value FROM result WHERE result.id = #TARGET_PARTITION#.OID';

DROP TABLE result;