help with calculating most likely period to increase

Hello, I am new and I really enter because I want to know how I can deal with this problem from knime

They ask me to calculate the period in which a person is most likely to increase their score by 50 points
the base I have for the exercise is from 2019 to 2040, and what they ask me is every how many years there is a significant increase in the score
The first thing I thought was to do a subtraction of all the years with the base month 2019 and categorize them into 1 and 0, however I feel that it is not the most optimal way to do this type of exercise, the other thing was to insert and use the base month 2019 , 2020, 2021… and so on for each period what happened is that I really don’t know how to do that hahaha and I don’t know if it’s correct, I just want to read opinions or if someone wants to help me I attach the base of the exercise

TESTMODEL1.xlsx (67.3 KB)

Thank you

Hello Melilla,

welcome to the forum. :slight_smile:
Let me repeat your request, so I can understand if I got it in the right way: you have a dataset containing scores for nearly 500 people starting from year 2019 to 2040. You need to calculate the average time in which a person increases his/her score by 50. This means calculating:

  1. when there is an increase in score of 50 points year by year for each person
  2. the difference in time for each user between two increases by 50 points
  3. the average difference between two increases, simply averaging all differences in time computed at point 2.

If yes, let’s proceed. If something is incorrect, tell me.

Have a nice evening,
RB

hello that’s right, I would just like to add what is the average time of all the people in which their score varies 50 points

So far what I did was think about the period in which its variance was greater
The other thing was to subtract all the periods with 2019 and if this increase by more than 50 points, assign it a 1 if not a zero and at the end add it, I think they are not the best ways to do it, that’s why I asked for a guide in the forum experts like you

Hi Mellilla,

first thing I’d do is change the dataset just like I did in the image below. The Unpivoting node will do the job.

Then you need to create a loop so you can work with just one person per iteration. I’d use the Group Loop Start node.
Then identifying the +50 points years will be easy. Also, calculating the numbers of years between two +50 increases will be easier.

Do you want me to build a workflow, or you just need some hints? :slight_smile:

RB

It would be very helpful if you help me create it, I think my current level of knime is not enough for creating loops.

Thank you

Hi Melilla,

sorry for the late reply, yesterday it was a tough day at work.
No problem in helping you, you’ll see that loops are super easy once you know the nodes available.

Have you tried to follow the steps I posted in my previous message? Do they make sense to you or is there something you were not convinced of?

Pop a message in case of need :slight_smile:

RB

1 Like

Hello, I would like to see it, I still really don’t understand the idea very well, as I told you, it is not something that I know how to master

Thanks

No problem! :slight_smile:

Here is a workflow I have prepared. Write me if something is not as it should.

Have a nice day,
RB

1 Like