Can you have a multilevel table?
78 views (last 30 days)
Show older comments
I have a large table that has several groups of similar variables. I'd like to create levels of variables so I can group and access them easily- perhaps I need a struct or perhaps I am being lazy.
load patients
T1 = table(LastName,Gender,Age,Height,Weight,Systolic,Diastolic);
head(T1,3)
I can merge variables, but I seem to lose the nested variable names.
T2 = mergevars(T1,{'Systolic','Diastolic'},'NewVariableName','BloodPressure');
head(T2,3)
I'd like to be able to access the Systolic and Diastolic variable under BloodPressure like this:
T2.BloodPressure.Systolic
%but not this
T2.BloodPressure(:,1)
It's a lot harder for me to keep track of the index, and I am hoping I would be able to use Tab Completion.
Also, in the event there is more than one 'Blood Pressure', say I have BloodPressure1 and BloodPressure2 both with sub variables Systolic and DIastolic, I'd like to be able to get all the Systolics at once- something like
T2.{:}.Systolic
Hope that explains what I'm after, and hoping there's a way to get there. Thanks.
0 Comments
Accepted Answer
Voss
on 25 May 2022
You can use 'MergeAsTable',true in mergevars:
load patients
T1 = table(LastName,Gender,Age,Height,Weight,Systolic,Diastolic);
head(T1,3)
T2 = mergevars(T1,{'Systolic','Diastolic'}, ...
'NewVariableName','BloodPressure', ...
'MergeAsTable',true);
head(T2,3)
T2.BloodPressure
T2.BloodPressure.Systolic
4 Comments
Seth Furman
on 31 May 2022
Starting in R2022a you can use patterns to index into tables (as well as in a number of table methods that accept table variable indices).
load patients
T1 = table(LastName,Gender,Age,Height,Weight,Systolic,Diastolic);
T2 = head(mergevars(T1,{'Systolic','Diastolic'}, ...
'NewVariableName','BloodPressure', ...
'MergeAsTable',true),3);
T3 = renamevars(T2,{'Height','Weight'},{'Systolic','Diastolic'});
T3 = mergevars(T3,{'Systolic','Diastolic'}, ...
'NewVariableName','BloodPressure2', ...
'MergeAsTable',true)
bp = T3(:, "BloodPressure"+wildcardPattern)
bpSystolic = varfun(@(t) t(:, "Systolic"), bp)
bpSystolic.Properties.VariableNames = bp.Properties.VariableNames
More Answers (1)
Adam Danz
on 16 Jan 2023
Edited: Adam Danz
on 16 Jan 2023
An alternative to @Voss's execellent answer is to create the nested table(s) and then add the nested table(s) as a table variable to the main table (MATLAB R2018b or later).
load patients
BloodPressure = table(Systolic,Diastolic);
T1 = table(LastName,Gender,Age,Height,Weight, BloodPressure)
3 Comments
Adam Danz
on 6 Feb 2023
> I really like nested tables, but how can I run groupsummary on them?
Demo:
t3 = table(randi(2,5,1),rand(5,1),'VariableNames',{'a','b'});
t2 = table(rand(5,1),rand(5,1),t3,t3,'VariableNames',{'c','d','t3','t4'})
t1 = splitvars(t2)
groupsummary(t1,'t3_a', 'mean', 't3_b')
See Also
Categories
Find more on Logical in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!