From: Regina Henschel on 3 Jun 2010 12:46 Hi Niels, Niels L. Ellegaard schrieb: > > The following is probably a feature request, but maybe someone can help > me find a good work around. > > I would like to be able to fit a trend line to a sub interval of a > dataset. So lets say that I have to the following dataset. > > x= 1,2,3,4,5,6,7,8,9,10 > y= 1,1,1,1,3,5,7,9,9,9 > > Now I want to draw a scatter plot that contains all the 10 data points, > but the trend line in the scatter plot should only be fitted to > datapoints (x,y) for which x>=4 and x<=8. This feature is be really > useful in high school physics classes because a lot of the time a > mathematical model only fits to a part of the data. > > As an example a high school student may heat water in a water cooker and > use termometer connected to a PDA to measure the temperature as a > function of time. If the effect is constant then the temperature will > rise as a linear function of time until the water starts boiling. Then > the temperature curve will level of and be constant. In order to > estimate the heat capacity of water the student needs to fit a trend > line to the data points in time interval where the water has not yet > boiled. I would insert a third column "relevant values", which contains only those values, which shall be used for the trend line. Then create the chart from all three columns, but calculate the trend line only from the third column. kind regards Regina  To unsubscribe, email: discussunsubscribe(a)openoffice.org For additional commands, email: discusshelp(a)openoffice.org
From: Niels L. Ellegaard on 3 Jun 2010 14:27 Regina Henschel <rb.henschel(a)tonline.de> writes: > I would insert a third column "relevant values", which contains only > those values, which shall be used for the trend line. Then create the > chart from all three columns, but calculate the trend line only from > the third column. Thank you for the answer. That was an elegant solution. But my guess is that most physics teachers try to ignore this problem as long as possible. Perhaps that is the best approach for now :) Kind regards Niels  To unsubscribe, email: discussunsubscribe(a)openoffice.org For additional commands, email: discusshelp(a)openoffice.org

Pages: 1 Prev: [discuss] Fitting a trendline to a subinterval of a dataset. Next: information, idea 