Changes

* ( You can scroll through earlier commands with: Cmd+ Arrow up)* Close the Grace "Results viewer" window* Close the interpreter window In the GUI, select:* For models, only select: "No_Rex" and "M61"* R1 parameter optimisation to False, ans this '''on-resonance''' data does not use R1.* The number of Monte-Carlo simulations to "3". (The minimum number before relax refuse to run).* Let other things be standard, and click execute '''NOTE:'''<br>The number of Monte-Carlo simulations is set to '''3'''.<br>This means that relax will do the analysis, and the fitted parameters will be correct, but the standard error of the parameters will be wrong. The reason for this is, that the data "naturally"? for some spins contains measurements which are "doubtful". <br>When relax is performing Monte-Carlo simulations, all the data is first copied x times to the number of Monte-Carlo simulations, and each R2eff point on the graph is modified randomly by a gaussian noise with a width described by the associated errors. For spins which contains measurements which are "doubtful", relax will be spending "very long time" trying to fit a meaningful model. <br>This "time" is poorly spend on "bad data". Rather, one should first try to * analyse the data quickly* make the graphs* examine which spins should be deselected* and then rerun the analysis with a higher number of monte-carlo simulations After this comes the part with global fit/clustered analysis.* analyse the data again* Select which spins should be included in a global fit/clustered analysis* and then rerun with analysis