How To Fix RSAN_RT322 - Tolerance < 0.02: Correlation coefficients saved, not regression coeffs

SAP Error Message - Details

• Message type: E = Error

• Message class: RSAN_RT - Messages for Runtime Analysis

• Message number: 322

• Message text: Tolerance < 0.02: Correlation coefficients saved, not regression coeffs

MESS SHORT TEXT:
Tolerance < 0.02: Correlation coefficients saved, not regression coeffs

What causes this issue?

The tolerance for one of the regression coefficients is smaller than
0.02.

System Response

A tolerance of below 0.02 means strong collinearity and thereby high
standard errors for the regression coefficients.
For this reason, bivariate correlation coefficients (which are not
influenced by collinearity) are saved instead of the regression
coefficients.

How to fix this error?

The system issues an error message and will not allow you to continue with this transaction until the error is resolved.

Procedure for System Administrators

Error message extract from SAP system. Copyright SAP SE.

Smart SAP Assistant

• What is the cause and solution for SAP error message RSAN_RT322 - Tolerance < 0.02: Correlation coefficients saved, not regression coeffs ?

  The SAP error message RSAN_RT322 indicates that during the execution of a regression analysis in SAP BW (Business Warehouse), the tolerance for the correlation coefficients was set below the threshold of 0.02. As a result, the regression coefficients could not be saved, and only the correlation coefficients were retained.
  Cause:
  
  Low Tolerance Level: The error is triggered when the correlation coefficients of the variables in the regression analysis are too low, indicating that the variables may not have a strong linear relationship. The default tolerance level is set to 0.02, meaning that if the correlation coefficient is below this value, the regression coefficients will not be saved.
  Data Quality Issues: The underlying data may have issues such as insufficient variability, outliers, or noise, which can lead to low correlation between the independent and dependent variables.
  Model Specification: The model may not be correctly specified, meaning that the chosen independent variables may not be appropriate for predicting the dependent variable.
  
  Solution:
  
  Increase Tolerance Level: If appropriate for your analysis, you can consider increasing the tolerance level in the regression settings. However, this should be done with caution, as it may lead to the acceptance of less reliable models.
  Review Data Quality: Check the data for quality issues. Ensure that there are no outliers, missing values, or errors in the dataset that could affect the correlation.
  Re-evaluate Model Specification: Consider whether the independent variables used in the regression are appropriate. You may need to include additional variables or remove those that do not contribute meaningfully to the model.
  Transform Variables: Sometimes, transforming the variables (e.g., using logarithmic or polynomial transformations) can help improve the correlation and the overall fit of the model.
  Consult Documentation: Review the SAP BW documentation for regression analysis to understand the implications of the tolerance settings and how to adjust them appropriately.
  
  Related Information:
  
  Regression Analysis in SAP BW: This is a statistical method used to model the relationship between a dependent variable and one or more independent variables. In SAP BW, regression analysis can be performed using the Analysis Process Designer (APD) or other tools.
  Correlation Coefficient: This is a statistical measure that describes the strength and direction of a relationship between two variables. A value close to 1 or -1 indicates a strong relationship, while a value close to 0 indicates a weak relationship.
  SAP Notes and Community: Check SAP Notes for any specific patches or updates related to regression analysis. The SAP Community forums can also be a valuable resource for troubleshooting and finding solutions from other users who may have encountered similar issues.
  
  By addressing the underlying causes and following the suggested solutions, you should be able to resolve the RSAN_RT322 error and successfully save the regression coefficients.