# Weekly Challenge

Solve the challenge, share your solution and summit the ranks of our Community!
New content is available in Academy! You may need to clear your browser cache for an optimal viewing experience

## Challenge #140: Prove the Birthday Paradox!

Highlighted
Quasar

Started down a macro route and then just abandoned it for the brute force method. One I'll come back to when I have more time to work the iterative macro out.

Spoiler

Highlighted
Atom

75% @ 33

90% @ 41

Bolide

Thought it would run faster based on tests, but it worked!

Spoiler
My solution involved a combination of batch and iterative macros.

Highlighted
Alteryx

Here is my solution for this weeks challenge.

Spoiler
Went the way that one can configure how many trials they want to run, so basically I created a Batch Macro to be able to run multiple iterations of the actual process and then put this Batch macro into a regular macro where the user can directly influence the number of trials.

Final Workflow:

Regular Macro to influence number of Trials:

Batch Macro to run the actual process multiple times:

Highlighted
Meteor

Hmm...I know it's not perfect.

Spoiler
Highlighted
Alteryx Certified Partner

Done !

Highlighted
Meteoroid

The program is way too slow, but I completed it using nested iterative macros. Also in bigger groups the probability rises a little bit faster and the last few observations have probability equals to 1.

Highlighted
Alteryx

I went down the route of allowing the data to explode a little rather than the macro route.

To be above 75% chance I was getting either group size 32 or 33.

Above 90% was 41 and above.

Spoiler
Highlighted
Asteroid

Highlighted
Alteryx

Why do you think its slow? So my solution just ran 10000 Simulations in 1:11 (as I configured mine in the way that I just have to enter how many simulations I want to run), haveing an i7 with 16gb ram, so I did not feel that was slow. The recommended 1000 it runs in 8 seconds.

By the way, I gotta say, the more simulations you run, the nicer the curve looks like, so 10000 already gives me a very nice shaped curve with almost no breakouts in between.

Spoiler
See here the curve after I just ran 50000 simulations in 8 minutes

Cheers

Oliver