If you know about the model, which is linear in your case, then you can use the following formula to calculate Y values corresponding the X values you want. Let's say you want to calculate Y where X=3.5. Then you should use:
=3.5*LINEST(B2:B6,A2:A6)
where B2:B6 is the range for Y values, and A2:A6 is the range for X values.
Other functions that may be useful for you:
LOGEST(known_y's,known_x's,const,stats)
In regression analysis, calculates an exponential curve that fits your data and returns an array of values that describes the curve. Because this function returns an array of values, it must be entered as an array formula.
GROWTH(known_y's,known_x's,new_x's,const)
Calculates predicted exponential growth by using existing data. GROWTH returns the y-values for a series of new x-values that you specify by using existing x-values and y-values. You can also use the GROWTH worksheet function to fit an exponential curve to existing x-values and y-values.
FORECAST(x,known_y's,known_x's)
Calculates, or predicts, a future value by using existing values. The predicted value is a y-value for a given x-value. The known values are existing x-values and y-values, and the new value is predicted by using linear regression. You can use this function to predict future sales, inventory requirements, or consumer trends.
Adding a Power Trendline seems to have worked. tusham has kindly saved me the trouble of explaining why:
The power trendline is given by y=c*x^b. Believe it or not, you can use LINEST to calculate the parameters. Here's how.
Take the log of the above equation to get ln(y)=ln(c)+b*ln(x). This is a linear equation!
and
A log-log display of the data doesn't change the underlying analysis. If you plot a bunch of X,Y pairs and ask for a power trendline, you will get the same result irrespective of whether the plot shows either or both of the axes on a log scale.
Best Answer
If you're looking for a predicted DATE when the Number Updated reaches 660, you need to write some formulas.
I did this in a spreadsheet I was using to track Backup growth at a previous job. The solution does not involve the graph, it uses the FORECAST() function of Excel.
The syntax is:
In your case X equals 660, Yrange is the Date, and Xrange is the Number of Successful Updates.
So if your dates are in column A and the TOTAL number of successful updates is in column B, and you've got less than a thousand values entered already, you'll put the following formula in a spare cell somewhere:
In my experience I was unable to tell it to just use all column values (A:A) without it failing somehow, so I had to set some arbitrary limit (A1:A1000) for it to work.
Format the cell you put this formula in as Date and you'll get a prediction as to when your number of installs will equal the number of users.
Excel Tips was a great help in finding this information for me back when I was building the spreadsheet myself: http://excel.tips.net/T002573_Using_the_FORECAST_Function.html