Challenge #159: April ENcyptanalytics

The last time I posted a challenge (#155) it dealt with string analysis and decrypting a message from a noisy block of text.

People have been interested in how I came up with the data block for the challenge.

I decided this would be a good opportunity to let you be creative and show me!

The challenge this week is to create an ENCODER for a message that is compatible with the DECODER we built in challenge #155.

The easiest thing to do would be to simply generate some number of copies of the input message
(even just copying out the original message), but let us disqualify all such non-encodings!

challenge_159_start_file.yxzp

Parse

Join

Preparation

Advanced

Data Analysis

Data Preparation

Transform

Difficult

Accepted answers

estherb47

And remarkably easy to turn this into an app. Bonus points if you get the reference of the text.

challenge_159_encryptor_App.yxzp

All comments

cgoodman

Here's my solution. My first 1st post

Spoiler

So I decided to encode my message by using the CharToInt function and then using a random number generate add a random number to the converted character to integer.

Challenge 159.PNG

A snippet of my encoded message:

Challenge 159 Encoded Message.PNG

challenge_159_solution.yxmd

patrick_digan

Fun!

Spoiler

I just generated random data using ASCII characters 33-133 and then replaced the most popular +least popular characters with the intended letter. I added the least popular letter to avoid any ties.

challenge_159_start_file.yxzp

Alekh

Fun! Will look for a more efficient method later. I had to make my own list of characters which probably made it more complicated than necessary.

Spoiler

2019-04-09 12_07_12-Alteryx Designer x64 - 159..yxmd.png

159..yxzp

afv2688

Interesting challenge

Spoiler

challenge_159_end_file.yxmd

danilang

Fun Challenge @TerryT .

Encoded strings

Encoded strings.png

Spoiler

Dan

challenge_159_solution_DL.yxmd

NicoleJohnson

Me: This looks like a tricky one, but I'm sure I can get there using the skills I already have using the basic tools that I'm most familiar with.

Also me: ... Although this seems like a good time to try learning Python, on an extremely difficult challenge, even though I've never used it before... right?!?

Spoiler

I created an app for mine, in the end - allows the user to enter a string to be encoded, along with desired minimum number of rows to be included in the block of text (more may be added in order to get to the desired max number per character in each position, but this at least ensures a minimum number of rows).

Workflow then parses the original message, feeds parameters into the Python tool, and uses the numpy.random.randint() function to create an array based on integers 32-127 (which will get converted back to standard A-Z a-z 0-9 & common symbol characters in the latter part of the workflow). I count the max number of occurrences of any character in each position of the resulting randomized array, and then add that many occurrences + 1 of the original characters to the array to ensure that my desired characters are the ones that occur the most. I then pivot the data, use a RandInt formula to come up with a new random order per position, sort the data using the new random order, un-pivot the fields, and then concatenate by row.

The results will then output your Encoded Text Block as well as a copy of your Original Text for validation.

Extremely pleased that I eventually figured this one out... and jumping in the deep end appears to be the way to go with Python, because that was SUPER fun and now I want to try more #neverstoplearning #andnevertakeeasywayout

Cheers!!

challenge_159_NicoleJohnson.yxzp

pasccout

Here is my solution

challenge_159_my_solution.yxzp

cplewis90

I am a huge fan of this series! Thanks @TerryT! Next maybe we can tackle KRYPTOS with Alteryx!

Spoiler

I definitely struggled with this one initially (should never try and do these at the end of a work day). I could not figure out how I was going to break a tie if two characters had the same number of occurrences. I then decided just grabbing the top two. It is a bit of overkill, but this way I can ensure success of no ties.
Challenge 159.PNG

challenge_159_CL.yxmd

RichoBsJ

Thanks @TerryT nice challenge!! Right now, I feel like a spy in the Cold War haha

Hope you like my solution

Spoiler

Bonus Track...Madrid it's coming

challenge_159_richobsj.yxmd

JoshuaGostick

It would be interesting to see how else you can play around with the workflow to create different types of codes.

Spoiler

challenge_159_JG.yxzp

erikschaap12

Nice challenge

Spoiler

challenge_159_solution.yxmd

ch12345

I had a go at this trying to reduce the number of tools. I think I succeeded, but the macro version is far slower than the generate rows, even for only 10000 rows!

Spoiler

Although I do prevent selection of NULL characters

challenge_159_ch.yxzp

RolandSchubert

I started using the unicode character set, switched to Latin-1 and excluded Null values.

challenge_159_solution_rsc.yxzp

ggruccio

W23

1O3

12W

T234

1H34

12A4

123T

W23

1A3

12S

F23

1U3

12N

challenge_159_gg_finish.yxmd

T_Willins

That took some work to figure out! My approach is a little different in that I used a sample tool to ensure enough of each character would be present for decoding.

Spoiler

Workflow 159.GIF

challenge_159_solution.yxmd

Ladarthure

My solution, very fun challenge!

challenge_159_start_file.yxmd

kat

Fun one!

Challenge #159 message.PNG

Spoiler

Challenge #159.yxzp

estherb47

May have overcomplicated it a touch, but it works! Now to turn it into an app so users can type their own phrase......

Spoiler

challenge_159_EHB_completed.yxmd

estherb47

And remarkably easy to turn this into an app. Bonus points if you get the reference of the text.

challenge_159_encryptor_App.yxzp

AbiramiJothi

Spoiler

Made my Monday, Thanks for this one.

Sample Results.PNG

Best,

Abi

challenge_159_start_file.yxzp

philipmannering

Solution attached.

Spoiler

Workflow

Challenge 159 - April Encryptanalytics.yxzp

sprodger

my solution:

Spoiler

challenge_159_solution_SP.yxmd

TerryT

@estherb47, I had to look it up, but was definitely worth the giggle!

https://vimeo.com/107786640

Thanks!

kelly_gilbert

Wow, this one was super fun.

First, I started out with some investigation of the data from Challenge 155:

Spoiler

* Oddly enough, the A (capital A, code 65) did not occur at all in the challenge 155 data??

* For each position within the string, the correct character occurred in that position about 5x as frequently as any other character (for example, character 1 was the correct character about 5% of the time, while other characters occurred about 1% of the time)

My solution:

Spoiler

I built an app that receives a string and number of sample strings from the user:

The outer workflow checks to make sure the user has selected a number of strings that is at least 10x the length of the string.

If the number of strings is appropriate, then the macro runs. This is a separate macro 1) to prevent running the Simulation Sampling tool unless the number of strings passes the 10x test, and 2) to allow the user to dynamically input the number of iterations (strings).

The macro creates [number of strings] * [length of input string] random characters, from 32 to 130. If the random character is between 32 and 126, the related ASCII character is used. If the random character is between 127 and 130, then the correct character for the position is used (giving the correct character 4 additional "chances").

The Simulation Sampling tool seems slower vs. using Generate Rows to create the number of characters with a Formula tool to generate the random character.

Also, the Simulation Sampling tool doesn't allow a uniform discrete distribution (it only has the option for a uniform continuous distribution). I compensated by sampling between 32.0 and 130.999999 and rounding down to the nearest integer.

Finally, example output:

Spoiler

challenge_159_solution_klg.yxzp

Wouter

My solution:

Spoiler

challenge_159_solution_WK.yxmd

Verakso

I remember challenge (#155) and that was fun.

This one was tough though, so my solution ain't pretty, but I got to the result in the end

giphy (1).gif

Spoiler

So I hade the same idea as many others, to create some random data, and make sure, that the string we actually want to hide, is having the most occurrences.

Well, after some time back and forth, I ended up with this.

My WorkflowIt ain't pretty but it works.

After I made it, I looked at some of the other solutions, and then I was ashamed, so I didn't even bother to make this as an app 😢

But it was still fun after all 😳

Still Climbing

/Verakso

challenge_159_solution_verakso.yxzp

TerryT

Hi @Verakso,

You solved it! You shared it! We all learn!

I've gotta admit I've learned and improved my solution based on other solutions as well!

It's what makes Alteryx Community great!!!

Thanks!

Terry T

JosephSerpis

Interesting Challenge

challenge_159_Joe_Serpis.yxmd

jasperlch

Solution attached.

challenge_159_JL.yxmd

Kenda

Spoiler

Why do small things that cause a lot of grief like this always happen to me? I built out my whole workflow and ran it only to find out it failed when passing through the macro. So I redid basically everything only to realize the only reason it was failing was because I created my own Text Input file that I called Field1 when the macro was expecting a field called data. *facepalm*

challenge_159_start_file.yxmd

jamielaird

Here's my solution

Spoiler

I created a dictionary of ~800 UTF-8 characters and performed a random character substitution for half of the letters in the string. With 10 differently encoded versions of the original text it can be cracked easily using the decoder.

Screenshot 2019-06-23 at 17.13.21.png

challenge_159_solution_JL.yxzp

LordNeilLord

Spoiler

challenge_159_LNL.yxzp

RWvanLeeuwen

My 1000 record long input:

159 my input.jpg

And my workflow to get there:

Spoiler

challenge_159_start_file - RWvL.yxmd

SeanAdams

Very different kind of challenge - thank you @TerryT

Spoiler

Solution was to split into characters; then add a random offset on the char ID; but for 1 in 10 of these set the offset to 0; then use a mod function to bring this back to the 1-160 range.

Added the 155 challenge decoder as a macro to check this.

challenge 155 data.yxdb

Challenge_159_SeanAdams_Solution.yxmd

DecoderMacro.yxmc

Quick Links

This months top contributors

mceleavey 383

mbarone 337

Hollingsworth 335

LanisC 335

RithiS 280