Taking dummy column names and create new column

Hi everyone.

My problem is to match similar companies and group them together. I've tried fuzzy matching, but it will not work for my current use case. Hence, I am trying another method.

As I am new to alteryx, I'm having problem with creating a new column below (highlighted in yellow). I have already transposed my matrix into dummy variables that contain similar words in the 'company' column. Now, all I need to do is to identify the dummies with '1' and append its column name onto the new column.

I think it is a simple solution, but I'm really stuck. I've tried multi-field formula, but not sure how I can create a 'new column' with it...

Please help!! Thank you.

Common Use Cases

Fuzzy Match

Accepted answers

Elias_Nordlinder

Hello @alpro-23 ,

I created three different solutions for you below.

1. If you already have created the 1's before.

You can first transpose the data.
After this you can filter where the values only are 1 and you will get your desired output.

2. If you do not have the 1's yet but have the column headers.

You can first transpose the data.
Create a new flag column where 1 if the original column contains the header.

After this you filter your data where the flag is 1 and you get your desired output.

3. If you want to use RegEX and not use Flag/Value at all.

Alternatively if all your dummy data looks like below where the desired words are the words of your original column, then you can use Regex_Replace where you only take the first two words.

//Regards

Elias

creating company_replaced.yxmd

All comments

mceleavey

Hi @alpro-23 ,

can you post some data (mock data is fine) and an example of where you are and what you're trying to achieve?

Thanks,

Elias_Nordlinder

Hello @alpro-23 ,

I created three different solutions for you below.

1. If you already have created the 1's before.

You can first transpose the data.
After this you can filter where the values only are 1 and you will get your desired output.

2. If you do not have the 1's yet but have the column headers.

You can first transpose the data.
Create a new flag column where 1 if the original column contains the header.

After this you filter your data where the flag is 1 and you get your desired output.

3. If you want to use RegEX and not use Flag/Value at all.

Alternatively if all your dummy data looks like below where the desired words are the words of your original column, then you can use Regex_Replace where you only take the first two words.

//Regards

Elias

creating company_replaced.yxmd

alpro-23

Thanks so much. This definitely solves my issue- and I knew the answer would be really simple! 🙂

Quick Links

This months top contributors

atcodedog05 19458

Qiu 15866

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703