bring don observations (retain) , By group

[seancho]

Dear all,

I have a 2020 product list, with detailed information. I vertically joined a 2020 list with older lists (2019,2018 etc.), so I have all the products ever produced, not just 2020 current products.

My goal is to get one observation for one ID, with all information. 

2020 list file: 

 ID       detailed info1(have more)  _2020 (indicator variable to show the product is in 2020 list) 

A101   Germany                               1

B101   China                                    1             

older file : ex)2019

ID        2019

A101    1

C101    1

Combined list , sorted by ID (I have it to this step)

ID        detailed 1      2020 2019 2018 

A101   Germany          1

A101                                     1

A101                                              1

B101    China             1

B101                                    1

C101                                    1

C101                                              1

PROBLEM: 

I want to make it look like; (tried retain but failed) 

ID        detailed 1 -10  2020 2019 2018 

A101   Germany          1

A101   Germany          1        1

A101   Germany          1        1        1

Then For final dataset: 

then using (last.ID ?) so I have one ID obs with most information 

A101   Germany          1        1        1

B101   China                1       1

C101                                      1        1

data product_sort;
	set product;
	by ID;

	if first.ID then do;
	

	if detailed_info_1 ne . then temp_1 = detailed_info_1;
	else if detailed_info_1 = . then detailed_info_1 = temp1;
	retain temp1;

	end;



/*keep only last symbol */

run;

Thank you so much for your help in advance! 

[novinosrin · Accepted solution]

data have;
input ID    $    detailed1   $   _2020 _2019 _2018 ;
cards;
A101   Germany          1 . .

A101      .              .             1 .

A101        . . .                                      1
B101    China             1 . .

B101         . .                           1 .

C101       . .                             1 .

C101     . . .                                         1
;

data want;
update have(obs=0) have;
by id;
run;

[smantha]

You can use a proc summary and do a max or proc means with a max and get the result.

proc means data = have nway missing noprint;

class ID;

var _:;

output out=want(drop=_type_ _freq_) max=;

run;

[smantha]

proc means data = have nway missing noprint;

class ID detailed_1  ;

var _:;

output out=want(drop=_type_ _freq_) max=;

run;

[seancho]

Thank you very much!!!