Can a trendline in Numbers be extended to see predictions without adding data that does not exist?

You can add an extra column for predicting 6 months in the future and using the trendline equation for the formula in that new column. You should be very wary of the result of such a prediction!

Something like this:

D10=1.7667*B10-1.3961

select D10 and fill down as needed to project

If you notice in my first response... the top screenshot shows a plot with a trend line that shows the equation. I used the slope and offset in the formula that extend the "prediction"

you could use the forecast() or linest() function to create this prediction with out the trend line.

Here is another example:

I entered actual (but made up) data in columns A and B

then I calculate the slope and intercept of that data in column F

F3=INDEX(LINEST(B2:B6,A2:A6,TRUE,TRUE), 1)

F4=INDEX(LINEST(B2:B6,A2:A6,TRUE,TRUE), 2)

now in columns C and D I enter X values and calculate predicted Y values using the slope and intercepts:

D2=$F$3×C2+$F$4

select D2 and fill down as needed

Hi Badunit,

I am replying to you because the original question was in November 2013.

In Numbers 3.2, plot the actual data. Then add another value for a prediction in (this example) cell B6, value 10. The trend line extends.

Regards,

Ian.

EC,

What Wayne showed you is only extending the Slope of the trend line because he happened to use the linear fit option. Had he chosen one of the several other options, he would have been doing more than just extending the Slope when he entered the equation of the Trend Line into the new series.

I think it is important to note that it is a fine distinction to worry about the extention of a trend line unless you also worry about the space between the data points.

Jerry

I want to extend the slope of the trendline without using predictions. I want the trendline to show me the predictions.

But isn't that precisely what Wayne's example does? It uses the trendline equation derived from the existing data points to plot additional points on the trendline.

Thanks for the useful example, Wayne. I've often wondered how exactly to do this.

SG

I would like to do this without adding false data that I don't have yet.

Extending a trendline (extrapolation) is exactly that. It says "If this trend continues, we can expect that the [number of whatever] will be [this] in six month's time."

Treading on thin ice!

Regards,

Ian.

B2-B9 were the X data in Wayne's example. B11-B14 appear to be somewhat arbitrarily chosen X data points for the extended trend line. B10 was a copy of B9, the last X data point of the actual data. All were probably typed in manually or he used a simple formula. For the extended trend line you really only need two X data points, not all those that Wayne had. Assuming you are extending to the right, you need one that is the same as the last real X data point and you need one that is further to the right, to the point you want the line extended.

If your trend is a linear fit (straight line), you can use FORECAST to calculate the Y data for your two X data points.

you could double click your mouse on the Trendline and then go to the Options tab and enter values into the Forecast boxes to extend the Trendline forwards or backwards beyond your data range depending on what you want. It does exactly the same as Wayne's example, but doesn't plot any points on the line.

Hi Badunit,

Best part about it is it works with non-linear trend lines too.

Yes. Y=X^2

Thanks!

Regards,

Ian.

And with some natural variation in the measured Y values

Regards,

Ian.

OK. I do not know how to do what you want then. I am sorry I misunderstood your question.

How did you get your trendline points back in the cells? (Cells B2-B14)

That looks like Excel to me, not Numbers.