Regression Analysis

Performs linear, logarithmic, or power regression analysis of a data set comprising one dependent variable and multiple independent variables.

For example, a crop yield (dependent variable) may be related to rainfall, temperature conditions, sunshine, humidity, soil quality and more, all of them independent variables.

მონაცემები

დამოუკიდებელი ცვლად(ებ)-ის (X) შუალედი:

Enter a single range that contains multiple independent variable observations (along columns or rows). All X variable observations need to be entered adjacent to each other in the same table.

დამოკიდებული ცვლად(ებ)-ის (Y) შუალედი:

Enter the range that contains the dependent variable whose regression is to be calculated.

ორივე, X და Y შუალედებს, ჭდეები ექნება

Check to use the first line (or column) of the data sets as variable names in the output range.

შედეგები:

მიმართვა ზედა მარცხენა უჯრედზე შუალედისთვის, სადაც შედეგები გამოჩნდება.

გამოტანის რეგრესიის ტიპები

Set the regression type. Three types are available:

• Linear Regression: finds a linear function in the form of y = b + a₁.[x₁] + a₂.[x₂] + a₃.[x₃] ..., where a_i is the i-th slope, [x_i] is the i-th independent variable, and b is the intercept that best fits the data.

• Logarithmic regression: finds a logarithmic curve in the form of y = b + a₁.ln[x₁] + a₂.ln[x₂] + a₃.ln[x₃] ..., where a_i is the i-th coefficient, b is the intercept and ln[x_i] is the natural logarithm of the i-th independent variable, that best fits the data.

• Power regression: finds a power curve in the form of y = exp( b + a₁.ln[x₁] + a₂.ln[x₂] + a₃.ln[x₃] ...), where a_i is the i-th power, [x_i] is the i-th independent variable, and b is intercept that best fits the data.

პარამეტრები

ნდობის დონე

A numeric value between 0 and 1 (exclusive), default is 0.95. Calc uses this percentage to compute the corresponding confidence intervals for each of the estimates (namely the slopes and intercept).

ნაშთების გამოთვლა

Select whether to opt in or out of computing the residuals, which may be beneficial in cases where you are interested only in the slopes and intercept estimates and their statistics. The residuals give information on how far the actual data points deviate from the predicted data points, based on the regression model.

თავისუფალი წევრი ნულის ტოლია

Calculates the regression model using zero as the intercept, thus forcing the model to pass through the origin.