Average No of Days

[#1]

Hi,

I have the following data. I want to calculate average decisioning days. I think average would not be the better to calculate it. Which one would be better method to calculate. Is there any option to calculate average no of days with and without outliers. Which approach could be the best one. Is there any option to use normal distribution or something else.

ID	Decisioning Days
11111	0
11112	1
11113	2
11114	1
11115	2
11116	2
11117	2
11118	1
11119	2
11120	1
11121	1
11122	3
11123	3
11124	1
11125	1
11126	2
11127	30
11128	25
11129	24

[agilos_mla]

Hi,

Please Check this post:

http://community.qlik.com/message/440542#440542

Michael

[jonathandienst]

Hi

You could exclude outliers based on fractiles (aka percentiles). For example this would exclude the top 1% and bottom 1% of the values:

Avg(Aggr(If(DecisioningDays < Fractile(Total DecisioningDays, 0.01) Or DecisioningDays  > Fractile(Total DecisioningDays , 0.99), DecisioningDays), ID))

HTH

Jonathan

[#4]

Jonathan,

Thanks but it does not calculate.

[agilos_mla]

You don't need the aggr sice Fractile is not an aggregation function. Try :

Avg(If(DecisioningDays < Fractile(Total DecisioningDays, 0.01) Or DecisioningDays  > Fractile(Total DecisioningDays , 0.99), DecisioningDays))

HTH,

Michael

[#6]

Michael,

It's giving me 33 decisioing days but it should be 2 days. As I have the follwoing frequency: Majority of apps fall under 4 days.

FinalDecision WorkingDays	No of Apps	%
0	110	1.48%
1	1941	26.05%
2	2044	27.43%
3	1725	23.15%
4	980	13.15%
5	207	2.78%
6	106	1.42%
7	67	0.90%
8	41	0.55%
9	22	0.30%
10	20	0.27%
11	12	0.16%
12	15	0.20%
13	14	0.19%
14	13	0.17%
15	14	0.19%
16	19	0.25%
17	9	0.12%
18	6	0.08%
19	10	0.13%
20	6	0.08%
21	8	0.11%
22	4	0.05%
23	9	0.12%
24	2	0.03%
25	1	0.01%
26	2	0.03%
27	3	0.04%
28	2	0.03%
29	5	0.07%
31	4	0.05%
33	2	0.03%
34	2	0.03%
35	1	0.01%
36	1	0.01%
37	1	0.01%
38	1	0.01%
39	2	0.03%
40	1	0.01%
41	1	0.01%
42	2	0.03%
43	2	0.03%
44	2	0.03%
45	3	0.04%
46	2	0.03%
47	1	0.01%
48	1	0.01%
49	2	0.03%
50	2	0.03%
54	1	0.01%
57	1	0.01%
Total	7452	100.00%