Re: Evaluate code fragment without making new cell for it?
- To: mathgroup at smc.vnet.net
- Subject: [mg127186] Re: Evaluate code fragment without making new cell for it?
- From: "Scot T. Martin" <smartin at seas.harvard.edu>
- Date: Wed, 4 Jul 2012 03:35:21 -0400 (EDT)
- Delivered-to: firstname.lastname@example.org
Based on your example, what you really should have are 4 cells. You shouldn't try to load everything into a single cell. This approach, while convenient to a beginner, defeats many purposes and makes for difficult to understand Notebooks and code in the medium to long term. Your present approach of copy-and-pasting, perhaps into a new Notebook even, is highly recommended. On Mon, Jul 2, 2012 at 2:25 AM, Nasser M. Abbasi <nma at 12000.org> wrote: > Sometimes I wish the following feature existed in > Mathematica notebook interface, but I can't seem to > find a way to do it. > > Sometimes, (while trying things) I find that I want to > evaluate small code fragment which sits in a cell that > contains more code. > > Currently what I do is either copy that code fragment to > a new cell and evaluate that cell, and when done, delete > the cell, or split the original cell at places where I want > to isolate the code fragment, then rejoin the cells when > done. > > It will be nice, if there was a feature, where one can just > select, using the mouse, the code fragment, and then right-click > and say evaluate-selected. > > The output will go into a new output cell as normal. > > btw, the above is something one can do with that other > system M____B > > For example, assume we have this toy cell > > ---CELL----- > x = 16 > Sqrt[x] > y = 3 > Sqrt[y*x] > --------- > > And I just wanted to evaluate only > > x=16 > Sqrt[x] > > There is no easy way to do it without splitting the > cell or making new cell as I mentioned above. > > This feature should not be hard to implement by > the front end I would think. The front end can > make an 'internal hidden cell' in its own buffer of the > code selected and evaluate it and send the output back. > > Do others think such a feature will be useful? > > --Nasser >