|
.png)
1. Exponential. The exponential function y = aebx will be used. The number e (approximately 2.7182) is the base of natural logarithms. The values of a and b will be calculated to best fit the data.
2. Linear. The linear function y = mx + b will be used. The values of m and b will be calculated to best fit the data.
3. Logarithmic. The logarithmic function y = cLn(x) + b will be used. The values c and b will be calculated to best fit the data.
4. Power. The power function y = axb will be used. The values a and b will be calculated to best fit the data.
5. Polynomial. The logarithmic function y = cnxn+ cn-1xn-1...+ b will be used, n being the polynomial order. The values cn and b will be calculated to best fit the data.
6. Moving average. Each yavg value will be the average from yn to y(n - p + 1) values. Being n the y index and p the number of periods to consider, yavg = (yn + yn-1, yn-2,... yn - p + 1)/n . For example, in the series 3, 5, 7, 2, 4 with the number of periods equal to 3, the first value will be (3 + 5 + 7)/3=5, the second will be (5 + 7 + 2)/3 = 4.66, and the last will be (7 + 2 + 4)/3 = 4.33
7. Forecast. Trendline y values can be calculated backward or forward the number of periods specified using the corresponding formula.
8. Name. A descriptive name for the trendline can be automatically or manually assigned.
9. Intercept. The point of intersection between the trendline and the Y axis can be automatically calculated or specified.
10. Display equation. The equation with the corresponding constant values used for calculation of Y trendline values can be shown or hidden.
11. Display r-squared value. The R-squared value measures how well the trendline fit the data series. The closer this value is to 1.0, the better the fit of the trendline to the data is.
|