What would be my category variable in this dataset?

[melhaf]

Hi,

If my dataset looks like this, what would be my category data for this?  (this is just fiction)

I want to have (I have two dataset, I want to compare two periods)  stacked columns diagram and I'm using this code:

/* Merge the calculated datasets for each period */
data merged_data;
merge calc_period1 (rename=(ECL_stage1=ECL_stage1_1 ECL_stage2=ECL_stage2_1 ECL_stage3=ECL_stage3_1))
calc_period2 (rename=(ECL_stage1=ECL_stage1_2 ECL_stage2=ECL_stage2_2 ECL_stage3=ECL_stage3_2));
run;

/* Stacked column chart */
proc sgplot data=merged_data;
vbar category / response=ECL_stage1_1 group=period1 groupdisplay=stacked barwidth=0.4;
vbar category / response=ECL_stage2_1 group=period1 groupdisplay=stacked barwidth=0.4;
vbar category / response=ECL_stage3_1 group=period1 groupdisplay=stacked barwidth=0.4;

vbar category / response=ECL_stage1_2 group=period2 groupdisplay=stacked barwidth=0.4 barattrs=(color=lightgray);
vbar category / response=ECL_stage2_2 group=period2 groupdisplay=stacked barwidth=0.4 barattrs=(color=lightgray);
vbar category / response=ECL_stage3_2 group=period2 groupdisplay=stacked barwidth=0.4 barattrs=(color=lightgray);

yaxis label="ECL (in millions)";
xaxis discreteorder=data;
keylegend / location=inside position=topright across=1;
run;

but I don't know what what my category variable would be.

[melhaf]

I don't know how to edit my thread, but the whole looks like this

proc sql;

    create table calc_period1 as select 

        ('EAD [EUR]'n * 'PD [%]'n * 'LGD [%]'n) / 1000000 as ECL_stage1,

        ('EAD [EUR]'n * 'LTPD [%]'n * 'LGD [%]'n) / 1000000 as ECL_stage2,

        ('EAD [EUR]'n * 'LGD [%]'n) / 1000000 as ECL_stage3

    from period1;

quit;

proc sql;

    create table calc_period2 as select 

        ('EAD [EUR]'n * 'PD [%]'n * 'LGD [%]'n) / 1000000 as ECL_stage1,

        ('EAD [EUR]'n * 'LTPD [%]'n * 'LGD [%]'n) / 1000000 as ECL_stage2,

        ('EAD [EUR]'n * 'LGD [%]'n) / 1000000 as ECL_stage3

from period2;

quit;