Display complete pattern of missing values in panel data

[Marco Giansoldati]

Dear Members,

I am interested in obtaining the pattern of missing and non-missing observations for a set of variables in a panel. It is important that the outcome of the command provides the complete list of missings and non-missings from the very beginning of the time span till the very end.

So far I found the commands:

missings

, and

mvpatterns

,

which do show me only the pattern of a subset of years, in a similar manner to what

xtdes

does.

I would be extremely thankfu if you could kindly provide me with an insight on how to obtain my desired outcome.

Thanks

Marco

[Nick Cox]

Can you make this concrete: #panels, #times, #variables?

[Firmin Clairant]

Dear,
you can try mdesc.

[Marco Giansoldati]

Dear Prof. Cox,

Thank you very much for your kind message.

I am dealing with a panel data structure with 41 years, 130 countries, and a number of variables I look at for every country.

Many thanks Professor.

Do you think it is possible to have an outcome similar to xtdes but for all the years in my time span?

Marco

[Marco Giansoldati]

Firmin Clairant . Thank you but I do not know if and how many are the missings for each country each year.

[Marcos Almeida]

This article, by approaching MI methods in panel data, discusses the question. Hopefully that helps.

[Firmin Clairant]

you can try this code:

bysort country year: mdesc

replace if possible country by panelvar_id.

[Marco Giansoldati]

Many thanks Firmin Clairant and Marcos Almeida

[Nick Cox]

As yet another suggestion I note missingplot (SSC)

[Marco Giansoldati]

Thank you very much Prof. . I do really appreciate the kind help you all kindly provided me