Splus 4.0 
Note these directions are for an old version of SPlus
Using Splus 3.3 objects in Splus 4.0
In general, objects created by Splus 3.3 should be compatible with Splus 4.0.
- If you have created files using data.dump() in 3.3 and then data.restore() in 4.0 should work:
- data.dump(objects(), "c:\\temp\\dumpfile") in 3.3
- data.restore("c:\\temp\\dumpfile") in 4.0
This is the safest way to save your data and transfer it between systems.
- If you have a _data directory (or a library) from 3.3, you can use attach().For example attach("a:\\_data") will allow you to use the objects in the _data directory on a diskette in the a: drive. You can also
- Click on the search path in the left window of the object browser
- Click on the search path icon with the right mouse button in right window
- This will pop up a menu and choosing "Attach database" will pop up a window prompting you to give the name of the database containing the saved data objects (e.g. a:\\_data in the example above.)
- After you've attached a database, to get the objects to appear in the object browser you need to do a filtering of the database. The filtering window is generated by clicking on the right mouse button in white space of the right browser window and choosing "Filtering..." from the pop-up menu.
Splus 3.3 seems to be able to use most Splus 4.0 objects.
Some other issues that may arise in using Splus 4.0
- <Ctrl C>.has been replaced by <ESC> as the interrupt key.
- <Ctrl Ins>/<Shift Ins> have been replaced by <Ctrl C>/<Ctrl V> for Copy/Paste.
- graphsheet() has replaced win.graph(), although the latter still works. You no longer need to explicitly open a graphics device before using a plotting function.
- The Object Manager in 3.3 has been replaced by the Object Browser in 4.0.
- There is a new data editor in 4.0.
- The File > File/Import and File > File/Export menus are different (different underlying software).
- Modules such as S+SpatialStatistics, S+Wavelets, etc., do not yet work with S-PLUS 4.0. New versions will be issued shortly.
- Loops no longer have a meaningful return value. In the
past
j <- 0
x <- for(i in 1:10) j <- j + i
would result in x==55 (same as j) but in 4.0 x will be vector("missing",0) which doesn't even get printed (j will still have the value 55). - The value of source("file") is now vector("missing",0) (i.e., not meaningful) if the last expression in file is an assignment. This was to work around another bug.
lm: June 28, 1999
