Help with the merge table

[NhatHoang1]

Hello. I've been struggling with question b. help needed

. The following data sets contain BMI measurements for the baseline, 3 months, and 9 months, for the intervention and control group patients.

data int;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

101         int          34.5 33.8 31.1

102         int          30.7 29.6 27.3

103         int          32.0 30.7 27.5

105         int          34.6 33.5 33.3

106         int          37.6 34.9 30.1

107         int          30.2 29.2 28.7

108         int          37.2 36.0 35.1

109         int          35.2 33.3 33.7

111         int          32.3 31.4 29.4

;

data cnt;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

401         cnt          33.7 33.4 33.2

402         cnt          31.2 30.1 29.8

404         cnt          30.8 31.1 31.8

406         cnt          29.2 29.0 28.9

408         cnt          27.7 27.3 27.9

409         cnt          33.8 32.9 31.4

a) Concatenate the data sets and compute the change in BMI between the baseline and 9 months. Your output should look like this.

Obs	ID	GROUP	BMI_base	BMI_3mos	BMI_9mos	BMI_change
1	101	int	34.5	33.8	31.1	-3.4
2	102	int	30.7	29.6	27.3	-3.4
3	103	int	32.0	30.7	27.5	-4.5
4	105	int	34.6	33.5	33.3	-1.3
5	106	int	37.6	34.9	30.1	-7.5
6	107	int	30.2	29.2	28.7	-1.5
7	108	int	37.2	36.0	35.1	-2.1
8	109	int	35.2	33.3	33.7	-1.5
9	111	int	32.3	31.4	29.4	-2.9
10	401	cnt	33.7	33.4	33.2	-0.5
11	402	cnt	31.2	30.1	29.8	-1.4
12	404	cnt	30.8	31.1	31.8	1.0
13	406	cnt	29.2	29.0	28.9	-0.3
14	408	cnt	27.7	27.3	27.9	0.2
15	409	cnt	33.8	32.9	31.4	-2.4

b) Remove the intervention group patients whose BMI decreased by less than 3 points and also the control group patients whose BMI increased over time. Your output should look like this:

Obs	ID	GROUP	BMI_base	BMI_3mos	BMI_9mos
1	101	int	34.5	33.8	31.1
2	102	int	30.7	29.6	27.3
3	103	int	32.0	30.7	27.5
4	106	int	37.6	34.9	30.1
5	401	cnt	33.7	33.4	33.2
6	402	cnt	31.2	30.1	29.8
7	406	cnt	29.2	29.0	28.9
8	409	cnt	33.8	32.9	31.4

[Reeza]

Please show what you've tried on your assignment so far and we can help from there.

---

@NhatHoang1 wrote:

Hello. I've been struggling with question b. help needed

. The following data sets contain BMI measurements for the baseline, 3 months, and 9 months, for the intervention and control group patients.

data int;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

101         int          34.5 33.8 31.1

102         int          30.7 29.6 27.3

103         int          32.0 30.7 27.5

105         int          34.6 33.5 33.3

106         int          37.6 34.9 30.1

107         int          30.2 29.2 28.7

108         int          37.2 36.0 35.1

109         int          35.2 33.3 33.7

111         int          32.3 31.4 29.4

;

data cnt;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

401         cnt          33.7 33.4 33.2

402         cnt          31.2 30.1 29.8

404         cnt          30.8 31.1 31.8

406         cnt          29.2 29.0 28.9

408         cnt          27.7 27.3 27.9

409         cnt          33.8 32.9 31.4

a) Concatenate the data sets and compute the change in BMI between the baseline and 9 months. Your output should look like this.

Obs	ID	GROUP	BMI_base	BMI_3mos	BMI_9mos	BMI_change
1	101	int	34.5	33.8	31.1	-3.4
2	102	int	30.7	29.6	27.3	-3.4
3	103	int	32.0	30.7	27.5	-4.5
4	105	int	34.6	33.5	33.3	-1.3
5	106	int	37.6	34.9	30.1	-7.5
6	107	int	30.2	29.2	28.7	-1.5
7	108	int	37.2	36.0	35.1	-2.1
8	109	int	35.2	33.3	33.7	-1.5
9	111	int	32.3	31.4	29.4	-2.9
10	401	cnt	33.7	33.4	33.2	-0.5
11	402	cnt	31.2	30.1	29.8	-1.4
12	404	cnt	30.8	31.1	31.8	1.0
13	406	cnt	29.2	29.0	28.9	-0.3
14	408	cnt	27.7	27.3	27.9	0.2
15	409	cnt	33.8	32.9	31.4	-2.4

b) Remove the intervention group patients whose BMI decreased by less than 3 points and also the control group patients whose BMI increased over time. Your output should look like this:

Obs	ID	GROUP	BMI_base	BMI_3mos	BMI_9mos
1	101	int	34.5	33.8	31.1
2	102	int	30.7	29.6	27.3
3	103	int	32.0	30.7	27.5
4	106	int	37.6	34.9	30.1
5	401	cnt	33.7	33.4	33.2
6	402	cnt	31.2	30.1	29.8
7	406	cnt	29.2	29.0	28.9
8	409	cnt	33.8	32.9	31.4

---

[NhatHoang1]

these are what I came up so far, but it only delete the int patients, I don't know what to do next to delete both int and cnt to merge in one table.

data int;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

101   int   34.5 33.8 31.1

102   int   30.7 29.6 27.3

103   int   32.0 30.7 27.5

105   int   34.6 33.5 33.3

106   int   37.6 34.9 30.1

107   int   30.2 29.2 28.7

108   int   37.2 36.0 35.1

109   int   35.2 33.3 33.7

111   int   32.3 31.4 29.4

;

data cnt;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

401   cnt   33.7 33.4 33.2

402   cnt   31.2 30.1 29.8

404   cnt   30.8 31.1 31.8

406   cnt   29.2 29.0 28.9

408   cnt   27.7 27.3 27.9

409   cnt   33.8 32.9 31.4

;

data concat;

set int cnt;

BMI_change=BMI_9mos-BMI_base;

if BMI_change>-3 then delete;

run;

proc print noobs;

run;

[Reeza]

Please post your code in a code block. 

data concat;
set int cnt;
BMI_change=BMI_9mos-BMI_base;
if group = 'int' and BMI_change>-3 then delete;
else if group = 'cnt' and BMI_change ... then delete;
run;
 
proc print noobs;
run;

also the control group patients whose BMI increased over time.

You need to first add conditional logic, so that it does different logic for each group. Then you need to convert the criteria above to code, increased over time means what in terms of the difference??

I'll let you convert that logic to SAS code by filling in the ... in the code above. 

---

@NhatHoang1 wrote:

these are what I came up so far, but it only delete the int patients, I don't know what to do next to delete both int and cnt to merge in one table.

data int;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

101   int   34.5 33.8 31.1

102   int   30.7 29.6 27.3

103   int   32.0 30.7 27.5

105   int   34.6 33.5 33.3

106   int   37.6 34.9 30.1

107   int   30.2 29.2 28.7

108   int   37.2 36.0 35.1

109   int   35.2 33.3 33.7

111   int   32.3 31.4 29.4

;

data cnt;

input ID$ GROUP$ BMI_base BMI_3mos BMI_9mos;

cards;

401   cnt   33.7 33.4 33.2

402   cnt   31.2 30.1 29.8

404   cnt   30.8 31.1 31.8

406   cnt   29.2 29.0 28.9

408   cnt   27.7 27.3 27.9

409   cnt   33.8 32.9 31.4

;

data concat;

set int cnt;

BMI_change=BMI_9mos-BMI_base;

if BMI_change>-3 then delete;

run;

proc print noobs;

run;

---

[tarheel13]

data int_cnt;
	set int cnt;
	*derive new variable;
	bmi_change=bmi_9mos-bmi_base;
	
	if group='int' then do;
		if bmi_change >-3 and bmi_change <0 then delete;
	end;
	if group='cnt' then do;
		if bmi_change>0 then delete;
	end;
run;

proc print data=int_cnt;
run;

I think it would be better for you to put bmi_change < 0 for the first condition. In this data, you did not have any that were no change, but technically a decrease would have to be a negative value in bmi_change. Also, increase means bmi_change would be > 0. 0 would represent no change.