Did you ever want to take the values for error bars in a chart from cells in a spreadsheet? Then you are maybe one of those who voted for Issue 366. As you can see by the issue number this is quite an old one. And it has over 130 votes. So, if you are interested in this feature, there are good news: it is implemented, integrated into the DEV300 code-base and thus will make it into OpenOffice.org 3.0!

It was done in the CWS chart20, which was integrated into DEV300.m2 (as chart20_DEV300). So, if you know how to do a build yourself, you can do this for this master workspace or a newer one. I suppose, there will soon be DEV300 developer snapshot builds available on the OpenOffice.org Download Website.

Here is an example for a chart that has error bars at its data points which use their values from spreadsheet data. In this example both, the positive and negative values share the same range of cells, so they have the same extent in both directions. Of course these ranges can be different as well.

 
In addition to this long awaited feature, the automatically calculated error-bars now also support standard error (also known as standard deviation of the mean). So now, there are the error types:

• constant value (positive and negative)
• percentage (one parameter for both, positive and negative)
• standard error
  
• standard deviation
• variance
• error margin (one parameter)
• cell-range (one for positive and one for negative values)

Technically, it would be no problem to have separate parameters for positive and negative values for percentage values and error margins as well, but the ODF file format currently does not allow this.

While implementing this new feature, the dialog for error bar properties had to be adapted. In that course, it was also improved in general. You might have noticed that in OpenOffice.org 2.3 there is one dialog/tab page for statistic properties that contains error bars as well as regression curves (now called trend lines). This dialog was split up into two dialogs for the trendline equation feature. The trendline part was already improved there. Now, the second part was done by cleaning up the error bar dialog. This is how it looks now:

tags: chart openoffice.org

In the upcoming OpenOffice.org 2.4 release there will be some nice new features in the Chart. And here they are in an overview. You can get more detail on the Wiki page about Chart Features for OOo 2.4. Thanks to Tony Galmiche there are nice screen shots of the new features together with the corresponding dialogs to access them.

If you want to try out the new stuff you can download a pre-release version. See this announcement for more information about the latest developer snapshot.

Equation and Value of R² for Trend Lines Now, it is possible to display the equation for a trend line (regression curve) next to it. The formula object can be moved around, formatted with a number format, font and graphical formatting. Alternatively, or in addition, the correlation coefficient R² can be displayed in the formula object. This has already been announced on GullFOSS. Wiki
Reverse Axes Axes can be reversed, i.e. they point in a different than usual direction. In the example the y-axis points down from the top, instead of the usual orientation. You can also reverse the category labels of a category axis. Wiki
Display Bars on Different Axes next to Each Other You can attach some bars of bar or column chart to the first axis, and some to the second one. As a result the bars are grouped for each axis and displayed one group before the other. Now, it is possible to arrange the bars side-by-side even when they are attached to different axes. In the example, the blue series is attached to the left axis, while the orange and red series are attached to the right axis. Wiki
Number Format for Data Labels You can set individual number formats for the value displayed in a data label, as well as for the percentage. (Wiki) Display both, the Value and the Percentage It is now also possible to display the value and the percentage at the same time. You can choose a separator like a comma, semicolon, space or newline. (Wiki)
Flexible Placement of Data Labels Data Labels can now be positioned at different anchor points of the data point objects. In the example you can see that in a bar chart there are the absolute positions above, below and center. In addition, it is also possible to set a position depending on the sign of the value. So, outside means below for negative values, and above for positive ones. Wiki
Best Fit Data Label Positioning for Pie Charts In addition to the before-mentioned positions for data labels, there is a best-fit algorithm for pie charts that tries to place labels without overlapping. Especially for smaller segments overlapping happens quite frequently, which can now be avoided. Wiki

tags: chart features openoffice.org

Issue 7998 is quite old, as you can see by the low number. It was submitted half a decade ago. It's issue: equations for trend-lines. This is one of the issues that haven't been done for quite a long time due to the re-implementation of the chart. Now, that we have the new chart in OpenOffice.org 2.3, I took care of it and finished almost by now. The UI is still not optimal. As we have a nice code-base now, the implementation was for sure much more fun than it would have been for the old implementation, but it was far more than a check-box (that's what was stated at one point in the comments of issue 7998) — I can tell you that much!

This is an example using the new feature. I chose only linear regressions in this example, but of course this feature works for all supported regression types. I was looking for some data that can nicely be used for regression curves, and found some here. And only after I created the chart I noticed that the new feature is really useful, you can see:

• R² is very small, i.e. the linear regression is not a very good match, or in other words the variance of the data is quite large.
• Ignoring the first fact you could interpret the equation and find out that the temperature rose about two percent over the years. It's up to you to conclude global warming out of this data, though.
  

So, whenever chart17 is integrated, this feature will make it into OpenOffice.org.

What you can do now is, insert an equation for each regression curve (that's how trend-lines are still called in OOo). You can display the equation itself, the square of the correlation coefficient R, or both together. The text can be formatted with borders, a fill style and different fonts and font effects. The number format for all numbers appearing in the formula or in R² can be set. The equations are initially placed on the left and aligned with the top of the bounding box of the curve, which is not the most sophisticated place, but you will usually move the formula around anyway to have it at a place where it is well visible.

The necessary file format extensions for this feature have been integrated into the ODF 1.2 draft by the OASIS OpenDocument TC.

P.S.: For those who are wondering about the smooth rendering: this is a screen shot of an exported PDF.

tags: chart openoffice.org

tags: chart community openoffice.org

tags: chart community openoffice.org

In the OpenOffice.org Wiki there is a page about the Chart. It became more and more a mixture of miscellany. So, I decided to clean up this mess a bit. The page now contains four major sections:

• Helping with the Chart. For developers there are some links that may be helpful for starting coding. For people who like doing QA, there are some predefined IssueZilla queries.
• Features. Grouped by different releases and features that are not yet decided upon. This was mostly done be Tony Galmiche.
• Technical Articles. Some of the things that were on the main chart page are now moved to separate sections. Any additional articles that go into details of the chart, either design or implementation, should go to Chart/Article Name and get a link here. We should avoid to put too many content on the main page.
  
• People. Overview about who is involved in the development of the Chart. If anybody thinks he is missing, go ahead, do changes on the page or drop me a note.

The main focus of these pages is on development. So, the section Development should be the one to grow most. I hope you find this page clear and useful. Any suggestions for improvement of this page are welcome.

tags: chart openoffice.org wiki

As you might know by now, in OpenOffice.org 2.3 there will be a new chart module. Although its main focus was the replacement of an unmaintainable previous implementation, there are still a lot of new features (Thanks a lot to Tony Galmiche, who added some nice pictures to the otherwise less impressive looking list).

The first thing I want to show here is the "right-angled-axis" Feature. This means you get a 90 degrees angle between  the x-axis and the y-axis in the 2d projection plane. Sounds complicated, but is quite easy to understand, especially with a picture.

So, what this feature does, is it combines the accuray of 2d charts and the niceness of 3d charts. You are still able to read and compare the values nicely, but also have this impressions of some more real columns. What you also see here is the "simple" 3d look, where the shading is done flat, i.e. every plane gets one shade and not a gradient. The more "realistic" look might look cooler, but again, when it comes to readability, the simple look might be the better choice.

This chart is almost default. In the Wizard I clicked on 3d on the chart type page. And I changed the colors, because green bananas don't taste so good, as well as red kiwis. But the colors are all taken from the new twelve-color palette for charts.

The other cool thing we have now, thanks to Regina Henschel, are spline-curves that do not require sorted x-values. For a demonstration here are some funny curves I calculated with some nice formulas in Calc.

 On the left side you see B-splines, on the right side you have cubic splines. Nice aren't they? If you ask yourself why the charts are rendered so nicely, I must admit that I cheated a little here. PDF export, cropping and conversion to PNG with Apple's Preview, scaled down with Photoshop Elements. I hope we have a rendering engine that renders such pretty pictures directly some day.

 P.S.: You can find the document with which I created these charts here.

tags: chart openoffice.org

Finally, the New Chart made it into the master build!

There are a lot of improvements. For details see the What's New page.

If you like, download the developer snapshot build and give it a try! In case you discover any problems you can check this list of existing Issues. You are welcome to submit new issues you find.

Thanks a lot to all contributors. There was plenty of valuable feedback from testing the milestone builds, suggestions for improvement, articles for the online help and specification documents. Many thanks to Regina Henschel, Tony Galmiche, Pierre-André Galmes, Jörg Wartenberg and all the other people of the OOo community who supported the Sun Chart team.

Example Charts created with SRC680_m213. 

tags: chart openoffice.org snapshot

Here are more charts that show some details of the Voting for new Chart Colors.

This chart shows the number of votes per day. What you can see in this chart is that the Saturdays are good days to give a vote, especially the first one. Another peak is on Wednesday the 27th of March. I have no idea what this might have caused. I didn't notice any blog or announcement on that day or the day before.

This graph shows the ten most common top-level domains of the sender's mails. Well, it is not very unusual that .com, .net and .org are quite common. It doesn't say much about the location of the people who voted. From the domains that clearly denote a country (well at least the country of your mail provider, not necessarily the one you live in), Germany has the highest number of mails, second is Poland and third is the Czech Republic.

Finally, we have some data about the mail programs used to send the submission mail. "Mozilla Products" are Mozilla Mail and Thunderbird in different versions, and make the biggest part of detectable mail clients. Second is Evolution, and on third place are Microsoft Mailers, like Outlook or Outlook Express. Others are programs that are not used very often (in this voting). The number of unknown programs is quite big, so the distribution might be quite different in reality. But the chart is nice, isn't it?

tags: chart

proposal number 2!

Thanks a lot to all of you who gave their vote for better default colors for the chart! The total number of valid votes is 632, which is an amazing number (This is more than the number of votes for the location of the OOoCon 2007).

You can see the results here. The following chart also shows the distribution of the votes for the different color sets. It was created with the new chart using the new colors.

Color proposal #2 won with almost a third of all votes, so it is a clear winner. Number two is proposal #4 with almost 15% of all the votes, and on rank three is proposal #3 with a bit more than 10%.

The new default colors will be integrated into the new chart implementation. The plan is to have that available in OpenOffice.org 2.3.

We are aware, that the new palette will not fit in every user scenario, and not every user will like the colors equally much. So, we plan to implement multiple color schemes in the Tools/Options/Charts dialog, so that you can choose from a variety of different schemes, even ones you created yourself. For progress on this see Issue 74992. Though, this issue does not have the highest priority, so it will not be the first thing to do after integration.

Anyway, I think the new colors are a very good thing to have, as they are the first thing you notice when you create a chart. The charts look much better per default, which is especially important for  people who don't use the chart very often, or just want to create a chart quickly without worrying about colors.

tags: chart user-experience