----- Forwarded message from [email protected] ----- Date: Mon, 15 Apr 2013 03:59:03 -0400 From: [email protected] Reply-To: [email protected] Subject: Re: independent variable files and regression To: [email protected]
Here is an example of independent variable file in .NTS format: First line in bracket you can write everything you feel useful. Then write the data format and specify the kind of matrix. In the case of one variable you write: 1 [matrix rectangular] 10 [number of cases-specimens] L [if specimens are labelled-if they are after the second row you need to add the name of specimens] 1 [number of variables =this case only one] and 0 [label for missing values...you can have -999 for missing data]. Alternatively open you TPS file in tpsRelw and save Centroid size..that will save you a .NTS that you can open with notepad and modify. Below is a file example for typical allometric analyses made on 3 species. The independent variable is size but it could be anything you like. Note also that the order must be the same as your .TPS file: " Centroid size for 3 species of mammals labelled 1 3L 1 0 SpeciesA SpeciesB SpeciesC -5.97144968030966E-002 -1.12439227273219E-001 4.99751702130639E-002 Below same file with no labels: " Centroid size for 3 species of mammals non-labelled 1 3 1 0 -5.97144968030966E-002 -1.12439227273219E-001 4.99751702130639E-002 Below a file with 3 species and two variables " Centroid size & body weight for 3 species of mammals non-labelled 1 3 1 0 -5.97144968030966E-002 2.00E+001 -1.12439227273219E-001 1.20E+001 4.99751702130639E-002 3.00E+001 Good luck Carlo [email protected] ha scritto: > > ----- Forwarded message from "[email protected]" ----- > > Date: Wed, 3 Apr 2013 10:54:30 -0400 > From: "[email protected]" > Reply-To: "[email protected]" > Subject: independent variable files and regression > To: [email protected] > > Hello guys, > > I would like to do a "Regression" in tpsRegr and I don´t know how > to do that. I´m following the instructions of an online workbook > called "A Practical Companion to Geometric Morphometrics for > Biologists: Running analyses in freely-available software" by Miriam > Leah Zelditch, Donald L. Swiderski, and H. David Sheets. > > The following text passage is from that online workbook. Here they > tell me how to do a regression in tpsRegr, however, I have no Idea > what an "independent variable file" could be and how to construct one. > > "Regression in tpsRegr > To run tpsRegr, you need a shape data file and an independent > variable file. The shape data file can be in either *.tps or *.NTS > format, the independent variable must be in *.NTS format. The first > line of the independent variable file thus contains the control line > giving the matrix type (1), number of rows (which is your sample > size), number of columns (the number of variables in the file), and > the missing data option (0). If you labeled the independent > variable, add an L to the third (column) number, e.g. 1L. The > remainder of the file consists of the column(s) of values for your > independent variable. For studies of allometry, you can produce your > independent variable file within tpsRegr if you used the scale > factor in tpsDig. Go to the File menu, select Save and then select > centroid size. That produces a file of centroid sizes in *.NTS > format. If you want to do the analysis using log-centroid size > instead, open your *.NTS file in a spreadsheet and add a second > column, this one of log-transformed values. Just remember to change > the control line to a 2 for the number of columns and whenever you > have two or more variables in the file, it is useful to include > their labels, so make sure that the control line has 2L for the > number of columns. Below that control line, add a line with the two > labels in it, e.g. CS LCS (no punctuation goes in this line). Save > that as a text file with an *.NTS extension (you may have to change > the name of the file if your spreadsheet automatically adds *.txt to > all files that are saved as text files)." > > Is there anyone among you who could help me out with this? > > Thank you very much, > > Bernd > > ----- End forwarded message ----- > > ----- End forwarded message -----
