Graph overlayed by id combined per group

[Corine EL Habr]

Hello everyone,

I have panel data that look like this (this is a subsample, it contains more years, more groups and more id per group)

Year	Group	id	characteristic	value
2016	A	a1	life	2
2016	A	a2	nl	5
2016	A	a3	re	3,5
2017	A	a1	life	3
2017	A	a2	nl	2,5
2017	A	a3	re	4
2016	B	b1	life	1,5
2016	B	b2	life	2
2016	B	b3	re	3,6
2017	B	b1	life	2,5
2017	B	b2	life	3,6
2017	B	b3	re	4

I want to draw a graph where Y is the value and the X is the year and:
Each group is represented in a subgraph
In each group, overlaid plots if characteristic == "life"

I have tried using xtline, but it does not allow for by(), so I have noway of dividing into subgraphs per group.

Thank you in advance

[Hemanshu Kumar]

Not really sure what you're trying to get here. Does something like this suffice?

separate value, by(characteristic) veryshortlabel
twoway line value? Year, by(Group) xlabel(2016(1)2017) scheme(s1color) legend(row(1))

[Corine EL Habr]

Thank you Hemanshu Kumar for your response.
The code you provided gives me a graph with one line for each characteristic. What I need instead is 1 line for each id, conditional on teir characteristic. For example, I want 1 plot per id for ids that have the characteristic life, by (group). As you can see in the data, in group B, I have 2 ids that have the characteristic life so I want 2 lines in the subgraph of group B. i do not necessarily need all characteristics to be used in 1 graph.

[Hemanshu Kumar]

So then

separate value, by(id) veryshortlabel
twoway line value? Year if characteristic == "life", by(Group)  xlabel(2016(1)2017) scheme(s1color) legend(row(1))