Not sure why not. From the HELP:
LINEST_B() returns the aggregated b value (y-intercept) of a linear regression defined by the equation y=mx+b for a series of coordinates represented by paired numbers in the expressions given by the expressions x_value and y_value, iterated over the chart dimensions.
Linoy, the linest functions consume a table with two related columns, one showing 'x' values, the other 'y' values (if you think of a cartesian x/y graph). These functions then calculate the 'slope' resp. the 'y-interception' of a linear regression.
If the y-values in that source tables are cumulative numbers or not doesn't really matter for the linest function, that's part of your task as a developer to create / calculate the correct numbers to feed into the linest functions.
(There is no option or something like that you can enable with the linest functions to enable an automatic accumulation, if that's what you were thinking of, like in the line graph).
Yeah, i saw that but can you please look at the sample example attached? The manual calculated regression line matches the automatic one but b from y=mx+b seems to be equal to Linest_M+Linest_B:
3 + (- 0.67) = 2.3333
The b value from the automatic equation is wrong, it's giving y-intersection value at x=1 (first x dimensional value), not at x=0. The equation clearly doesn't match the shown line.