# Thread: Chart 2 series of different lengths? ((2000/SR-1a))

1. ## Chart 2 series of different lengths? ((2000/SR-1a))

I am charting two series of data, and I cannot figure out a way to have the second series, which starts a year later, start at the correct point. Instead, it displays as though the data belonged in an earlier period. The partial workaround seems to be to put zeros in for data for the earlier periods, but then the trend line is completely meaningless. Am I missing some obvious "start at" parameter?

2. ## Re: Chart 2 series of different lengths? ((2000/SR-1a))

Have you defined the x-axis values for both series correctly? What you want should be no problem. See attached workbook.

3. ## Re: Chart 2 series of different lengths? ((2000/SR-1a))

The columns I was using were not contiguous, so I defined the two series and the X axis separately. Moving the data into contiguous columns makes it easier, but the trend line doesn't seem to behave any better. Here's an example using a polynomial function:

4. ## Re: Chart 2 series of different lengths? ((2000/SR-1a))

Do you use the same X-values for your second series? Then you must have some missing data in the Y-series. The trend line then extrapolates through all the X-values in the plotted series. So, if you leave out the X-values that correspond to missing Y-values, the trend line will be what you want.

5. ## Re: Chart 2 series of different lengths? ((2000/SR-1a))

The X-values, the labels that appear along the X axis on the bottom, are the same for both series. It is the number of Y-values that is different: one is a 2.75 year set of data and the other is only a .75 year set. If I leave out the X-values for the previous two years, it is a different chart; doesn't tell the same story. Am I missing something?

If this wasn't updated every month, I'd capture the trend line as a graphic from a different chart and lay it on top, but that doesn't really help me in the long run.

(I tried to find a way to lay a second chart over the first, but found no way to give it a transparent fill...)

6. ## Re: Chart 2 series of different lengths? ((2000/SR-1a))

What I mean is that the second series contains X-values that do not correspond to Y-values (because they are missing; and they are missing because there is no Y-data for the 2 years before the .75 year period). If this series is then used to calculate a polynomial trend line, then it will use only the data of the .75 year to calculate the function, but when plotted, it will be plotted over the whole 2.75 year period, resulting in this odd behaviour for the 2 years preceding the .75 year period. This is due to the extrapolation (or forecasting) over a previous period.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•