Dear all,
I would like to produce a panel plot of three scatterplots; these scatterplots should be arranged horizontally. The problem is that while there is a common variable for the y-axis (‘dep_var’) there are three separated independent variables ('ind_var1', 'ind_var_2', 'ind_var3'). So unlike the standard situation where the relation between y and x is grouped by a single categorical variable, in my situation I have three separate scatterplots. Is it still possible to arrange them using syntax in a horizontal output? Many thanks for your suggestions - N. -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
You need to create the "standard situation". Reformat data to long format and
use the index to split the ind_vars to panels. HTH, PR *Demo data. DATA LIST list / id iv1 iv2 iv3 dp (5F8.1). BEGIN DATA 1 1 2 1 1 2 2 1 3 2 3 3 3 2 3 END DATA. *don't play with main dataset. DATASET COPY Test. DATASET ACTIVATE Test. *keep relevant vars. MATCH FILES File * /KEEP dp iv1 iv2 iv3 . *to long format. VARSTOCASES /MAKE iv FROM iv1 iv2 iv3 /INDEX p 'Panel' (3) . *Name the panels. VALUE LABELS p 1'iv1' 2'iv2' 3'iv3'. * Do GGRAPH with iv and dp with p as panel varaiable. GGRAPH /GRAPHDATASET NAME="graphdataset" VARIABLES= iv dp p MISSING=LISTWISE REPORTMISSING=NO /GRAPHSPEC SOURCE=INLINE. BEGIN GPL SOURCE: s=userSource(id("graphdataset")) DATA: iv=col(source(s), name("iv")) DATA: dp=col(source(s), name("dp")) DATA: p=col(source(s), name("p"), unit.category()) GUIDE: axis(dim(1), label("iv")) GUIDE: axis(dim(2), label("dp")) GUIDE: axis(dim(3), label("Panel"), opposite()) SCALE: cat(dim(3), include("1", "2", "3")) ELEMENT: line(position(iv*dp*p), missing.wings()) END GPL. *clean up. DATASET CLOSE Test. nina wrote > Dear all, > > I would like to produce a panel plot of three scatterplots; these > scatterplots should be arranged horizontally. > > The problem is that while there is a common variable for the y-axis > (‘dep_var’) there are three separated independent variables ('ind_var1', > 'ind_var_2', 'ind_var3'). So unlike the standard situation where the > relation between y and x is grouped by a single categorical variable, in > my > situation I have three separate scatterplots. > > Is it still possible to arrange them using syntax in a horizontal output? > > Many thanks for your suggestions - > N. > > > > -- > Sent from: http://spssx-discussion.1045642.n5.nabble.com/ > > ===================== > To manage your subscription to SPSSX-L, send a message to > LISTSERV@.UGA > (not to SPSSX-L), with no body text except the > command. To leave the list, send the command > SIGNOFF SPSSX-L > For a list of commands to manage subscriptions, send the command > INFO REFCARD -- Sent from: http://spssx-discussion.1045642.n5.nabble.com/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
In reply to this post by nina
The STATS REGRESS PLOT extension command, which can be installed from the Extensions > Extension Hub menu if not already present, can do horizontally arranged plots of a continuous y variable against a number of continuous or categorical variables. For categorical X variables, the charts can be bar charts, line charts, or boxplots. A scatterplot would be dubious as it implies a continuous X axis. There is a single Y axis across the charts for each horizontal set. You can control the size of each panel and other aspects. The command generates and executes GGRAPH/GPL syntax to carry out the plots. On Tue, Nov 27, 2018 at 12:17 AM nina <[hidden email]> wrote: Dear all, |
Free forum by Nabble | Edit this page |