topic Re: WEIBULL distribution in QlikView

WEIBULL distribution

Cvarela13 — Mon, 16 Nov 2020 16:19:02 GMT

I have a requirement in qlikview, for which one of the steps is to execute a weibull distribution, I know it does not exist in qlikview, but I would like to know if anyone knows if there is any way to apply the weibull distribution

Re: WEIBULL distribution

chrismarlow — Mon, 16 Nov 2020 16:59:38 GMT

Hi,

When you say execute a weibull distribution do you mean;

1) Draw a distribution given (static) shape and scale, or,

2) Best fit a weibull distribution to a set of data (i.e. calculate what the shape and scale should be somewhat dynamically)?

Cheers,

Chris.

Re: WEIBULL distribution

Cvarela13 — Mon, 16 Nov 2020 17:17:48 GMT

Hi,

I think it looks more like the second option, it would be to replicate the formula in script like in Excel to a data set in a table:

WEIBULL (x, alpha, beta, cumulative) = Result

https://www.excelfunctions.net/excel-weibull-function.html

Re: WEIBULL distribution

chrismarlow — Tue, 17 Nov 2020 10:05:47 GMT

Hi,

So from here;

https://www.real-statistics.com/distribution-fitting/fitting-weibull-regression/

And a spreadsheet I found here;

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwju0uuRpIntAhXKesAKHfw-CxEQFjABegQIBhAC&url=https%3A%2F%2Fwww.researchgate.net%2Fprofile%2FPaul_Louangrath%2Fpost%2FIs_there_any_other_method_of_measuring_the_performance_of_a_mutual_fund_scheme_than_Sharpe_Ratio_Beta_and_R_Squared%2Fattachment%2F59d625f4c49f478072e9a905%2FAS%253A272168477233166%25401441901429689%2Fdownload%2FWeibull-Analysis-In-Excel.xls&usg=AOvVaw0r53VoOiTuJUIRjaxcQC7D

On a toy application based on the spreadsheet I can generate the following with standard chart functions;

So it appears that the inputs to Weibull CDF/PDF can be calculated, from there these could possibly be fed into the long form of the CDF/PDF to plot the fitted distribution vs the actual (possibly using a variable to simplify?).

If this looks like it is going in the right direction happy to try to help further, a couple of caveats;

1) Examples I've seen overlaying a normal distribution have generally relied on CLASS calculated dimension to group the values & don't look great where the values are quite sparse (there may be a way round that though)

2) The stats here is we(ibu)ll out of my comfort zone ... so someone competent will need to sense check this is not a load of (wei)bull 😀

Cheers,

Chris.

Re: WEIBULL distribution

chrismarlow — Tue, 17 Nov 2020 10:21:39 GMT

Hi,

One thing that did occur after posting this is that there are ways of getting interoperability between R & QlikView, link below may be a starting point;

https://community.qlik.com/t5/QlikView-Integration/Usage-of-R-programming-with-Qlikview/td-p/1014803

This may or may not be simpler than the approach above in this case, but might open up more that you are interested in for the longer term.

Cheers,

Chris.