Do you have a simplified example with a code? It is not clear to me what do you mean by tree but if you refer to tree data structure, maybe you could change the data structure to tree (https://cran.r-project.org/web/packages/data.tree/vignettes/data.tree.html) and try to write comparison of two tree objects. It might be easier that data.frame alone.
On 12 September 2017 at 12:27, Céline Lüscher <c-luesc...@hispeed.ch> wrote: > Hi everyone, I need your help to solve a problem with occurrence and two > dataframes. > I have an excel table of 15200 lines. Each line correspond to a tree analyzed > for its structures. I have all the structures in columns (48 structures). The > occurrence of these structures has been counted on every tree. For example, > the tree 12607 has 3 structures CV11, 1 structure IN12 and none (0) of the > rest of all the other structures. The very last column is the value given to > the tree, according to the structures found on it (each structure giving a > number of point to the tree by its presence on it). > The question is: Are there some structures, or combination of structures, > which give a high value to the tree ? Of course, according to the value of > each structure, we can see which one has a higher value than the others (ex: > structure CV11 has a value of 15, structure IN12 has a value of 4). But what > I want to know is, if we take all the trees having a final value higher than > 100 (we create a new dataframe "data100"), and we compare with the trees > having a final value under 100 (we create another dataframe "data0"), can we > find a significant difference in the number and occurrence of structures > found on these trees? And which structure is related to trees with a higher > value than 100 ? > For now, I have only a visual answer to the question. I did two boxplot of > the data100 and data0, and I have seen some différences : 2 structures are > only found in the data100, which can be caracteristic of a final value higher > than 100. The problem is that I’m looking for a test to prove this. > If you have any idea or proposition for solving this problem.. it will be > great! > Best wishes, > C. > > > Gesendet von Mail für Windows 10 > > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.