Nested Aggr - Set Analysis

[phoenix]

I need to get the top 10 highest Due % products (only products with balance >$25M) in the portfolio for a specific month.

I am able to get to the top 10 highest Due % products (for all products) using 

=if(aggr(rank(sum({<Date={20181231}>}Due)/sum({<Date={20181231}>}UNPD_PRIN_BAL_AMT)),Product)<=10,Product)

but cannot figure out how to do the aggr(rank()) on the subset of products with balance >$25M for that specific month

Products with Balance > 25M
=if(aggr(sum({<Date={20181231}>}Balance),Product)>25000000,Product)	Due%	Balance
A	1.08%	$1,307,394,657.97
B	1.21%	$744,059,134.25
C	0.77%	$357,608,576.29
D	0.64%	$254,258,526.58
E	0.72%	$225,546,921.52
F	0.70%	$214,473,494.20
G	1.25%	$208,853,674.18
H	0.73%	$140,247,509.05
I	0.90%	$136,058,009.82
J	0.96%	$104,991,278.33
K	0.59%	$99,152,698.37
L	1.95%	$93,547,731.32
M	0.60%	$63,879,678.08
N	2.05%	$49,213,667.16
O	0.54%	$48,322,095.78
P	1.46%	$40,884,035.28
Q	0.83%	$39,308,599.38
R	1.03%	$38,115,956.84
S	0.49%	$34,866,485.54
T	0.63%	$33,443,292.79
U	0.32%	$26,726,725.36
Top 10 due % Products(Balance > 25M)
?????	Due%	Balance
N	2.05%	$49,213,667.16
L	1.95%	$93,547,731.32
P	1.46%	$40,884,035.28
G	1.25%	$208,853,674.18
B	1.21%	$744,059,134.25
A	1.08%	$1,307,394,657.97
R	1.03%	$38,115,956.84
J	0.96%	$104,991,278.33
I	0.90%	$136,058,009.82
Q	0.83%	$39,308,599.38

[bgerchikov]

Try this one for dimension:

=if(aggr(sum({<Date={20181231}>}Balance),Product)>25000000 And aggr(rank(TOTAL sum({<Date={20181231}>}Due)/sum({<Date={20181231}>}UNPD_PRIN_BAL_AMT)),Product)<= 10,Product)

Good Luck!

[phoenix]

tried this but it is doing aggr(rank(TOTAL sum({<Date={20181231}>}Due)/sum({<Date={20181231}>}UNPD_PRIN_BAL_AMT)),Product)<= 10 and then filtering Products for >$25M. i want it to filter the products with $25M first and then do the Due % rank

[bgerchikov · Accepted solution]

I see.

Here is another option that works for me:

=if(aggr(rank(TOTAL sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>25000000"}>}Due)/sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>25000000"}>}UNPD_PRIN_BAL_AMT)),Product)<= 10,Product)

[phoenix]

I thought it is working but I am getting different results between

=if(aggr(rank(sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Due)/sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Balance)),Product)<= 25,Product)

and

=if(aggr( sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Due ) /sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Balance ) ,Product )>aggr( sum(total {< Date={20181231} >} Due ) /sum(total {< Date={20181231} >} Balance ) ,Product ) ,Product ) 

 highlighted in BOLD are additional results I am getting from 2nd set analysis which i dont see in the first one

if(aggr(rank(sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Due)/sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Balance)),Product)<= 25,Product)	Due %	Balance	if(aggr( sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Due ) /sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Balance ) ,Product ) > aggr( sum(total {< Date={20181231} >} Due ) /sum(total {< Date={20181231} >} Balance ) ,Product ) ,Product )	Due %	Balance
LM	3.64%	$1,050,053.35	LM	3.64%	$1,050,053.35
Fr	2.53%	$1,581,308.10	Fr	2.53%	$1,581,308.10
Int	2.20%	$1,105,440.39	Int	2.20%	$1,105,440.39
Ma	2.13%	$3,310,626.70	Ma	2.13%	$3,310,626.70
Ti	2.12%	$1,246,107.42	Ti	2.12%	$1,246,107.42
De	2.10%	$1,249,319.67	De	2.10%	$1,249,319.67
Te	2.09%	$7,297,799.99	Te	2.09%	$7,297,799.99
W	2.06%	$3,703,667.27	W	2.06%	$3,703,667.27
Big	2.05%	$49,213,667.16	Big	2.05%	$49,213,667.16
Ge	2.01%	$1,648,498.29	Ge	2.01%	$1,648,498.29
Su	1.95%	$93,547,731.32	Su	1.95%	$93,547,731.32
CMT	1.88%	$4,924,117.17	CMT	1.88%	$4,924,117.17
Lo	1.70%	$16,355,766.39	Red	1.70%	$1,550,989.61
Ca	1.69%	$2,557,714.08	Lo	1.70%	$16,355,766.39
Lo	1.68%	$1,301,617.45	Ca	1.69%	$2,557,714.08
Ga	1.54%	$4,549,460.06	Lo	1.68%	$1,301,617.45
Tu	1.53%	$1,872,888.40	Ga	1.54%	$4,549,460.06
Ho	1.50%	$4,225,626.65	Bo	1.54%	$8,474,888.61
Br	1.45%	$2,353,437.39	Tu	1.53%	$1,872,888.40
Vi	1.45%	$1,317,374.95	Bu	1.52%	$1,997,419.22
AB	1.44%	$1,085,893.87	Ho	1.50%	$4,225,626.65
St	1.40%	$3,181,642.72	Br	1.48%	$7,890,247.90
Lak	1.37%	$3,970,718.85	PJ	1.46%	$40,884,035.28
En	1.35%	$1,941,699.58	Bro	1.45%	$2,353,437.39
Ap	1.29%	$4,901,928.38	Vi	1.45%	$1,317,374.95

[phoenix]

Set Analysis 1:

=if(aggr(rank(sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Due)/sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Balance)),Product)<= 25,Product)

Set Analysis 2:

=if(aggr( sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Due ) /sum( {< Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"} >} Balance ) ,Product )>aggr( sum(total {< Date={20181231} >} Due ) /sum(total {< Date={20181231} >} Balance ) ,Product ) ,Product ) 

Set Analysis 3: (two dimensions)

=Product

=if(Date=20181231,Date)

Set Analysis 4:

=if(aggr(rank(sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Due)/sum({<Date={20181231},Product={"=sum({<Date={20181231}>}Balance)>1000000"}>}Balance)),Product,Date)<= 25,Product)

Out of these Four straight tables I am seeing more results from set Analysis 3,4 and they both match. Not sure why first two are not working. I can use Set analysis 4, but rank is not working in that. It is giving me all results for (>$1M) instead of top 25. how to make Rank function work in the Set analysis 4? Thanks

[bgerchikov]

Hi Phoenix,

Not sure about your second expression. For example, Product "Lo" appears 2 times. 

Can you upload your app?

Sorry, I'm away from my testing environment right now....

[bgerchikov]

OK. Let's try to go in different direction.

For dimension you will use your initial Aggr statement, which limit your products by 25M or so.

In chart expression you can use the rank function

=if(rank(TOTAL sum({<Date={20181231}>}Due)/sum({<Date={20181231}>}UNPD_PRIN_BAL_AMT)),Product)<= 10,sum({<Date={20181231}>}Due)/sum({<Date={20181231}>}UNPD_PRIN_BAL_AMT)), Null())