splitapply table for zscore

I have a table in the format:

Group Score
A 3
B 7
C 18
A 10

How do I combine the splitapply function for table variables with zscore , such that it calculates zscore(Score) according to Group rather than the whole table?

I have tried:

[G,results] = findgroups(Table.Group);
ZS = splitapply(@zscore,Table.Score,G);
results.ZS = ZS

Answer 1

[Peter Perkins]

splitapply is more for computing scalar summaries. Try using varfun, with a grouping variable, returning a table. Something like

t2 = varfun(@zscore,t,'GroupingVariable','Group')

[UQFG]

Thanks, that does seem to be appropriate... however, when I try to apply it, I get the following errors:

Error using getGroupingVarOrTime (line 13) Unrecognized variable name 'Table.Group'.

Error in tabular/varfun (line 130) groupVars = getGroupingVarOrTime(groupVars, a);

[Peter Perkins]

I'm guessing you have a table named 'Table' (which is probably not a great idea), and a variable in it named 'Group'. The error is telling you that you do not have a variable named 'Table.Group'.

[UQFG]

I actually don't... I just used those names as generic examples!

I just realised that you don't need the "Table." in front of the Group name... so the code that works is:

ZS=varfun(@zscore,Table,'GroupingVariables','Group')

Whereas what I had before was:

ZS=varfun(@zscore,Table,'GroupingVariables','Table.Group')