Manuals/calci/REGRESSIONANALYSIS

Revision as of 15:20, 20 December 2016 by Devika (talk | contribs)
REGRESSIONANALYSIS (YRange,XRange,ConfidenceLevel,NewTableFlag)


  • is the set of dependent variables .
  • is the set of independent variables.
  • level of Confidence value.
  • is either 0 or 1.

Description

  • This function is calculating the Regression analysis of the given data.
  • This analysis is very useful for the analyzing the large amounts of data and making predictions.
  • Regression analysis is a form of predictive modelling technique which investigates the relationship between a dependent and independent variable.
  • This technique is used for forecasting, time series modelling and finding the causal effect relationship between the variables.
  • This analysis give the result in three table values.
  1. Regression statistics table.
  2. ANOVA table.
  3. Residual output.

1.Regression statistics :

  • It contains multiple R, R Square, Adjusted R Square, Standard Error and observations.
  • R square gives the fitness of the data with the regression line.
  • That value is closer to 1 is the better the regression line fits the data.
  • Standard Error refers to the estimated standard deviation of the error term. It is called the standard error of the regression.

2.ANOVA table:

  • ANOVA is the analysis of variance.
  • This table splits in to two components which is Residual and Regression.
Total sum of squares = Residual (error) sum of squares + Regression (explained) sum of squares. 
  • Also this table gives the probability, T stat, significance of F and P.
  • When the significance of F is < 0.05, then the result for the given data is statistically significant.
  • When the significance of F is > 0.05, then better to stop using this set of independent variables.
  • Then remove a variable with a high P-value and return the regression until Significance F drops below 0.05.
  • So the Significance of P value should be <0.05.
  • This table containing the regression coefficient values also.

3.Residual output:

  • The residuals show you how far away the actual data points are from the predicted data points.