ios-personmd-notifications md-help-circle

Profile

  • Guest
    medal 0
  • Posts: 21
  • Post Likes: 3765

Notifications

  • No Unread Notifications

Unresolved
CSV and data analysis

warning
This thread is closed. Threads older than 6 weeks are closed automatically. To continue this discussion, create a new thread.
angle-double-left ios-arrow-back 1 ios-arrow-forward angle-double-right
medal 5000
4 years 355 days ago
Hi! 
Do you know any tools to analyse the data from CSV file? 
I tried to do this with old good Excel but there is no appropriate data format for lap times. 
Could someone help with this case? 
md-quotelink
medal 5000
4 years 354 days ago (Last edited by Wambou Wambou 4 years 354 days ago)
Hi

I did not try to open races' data as CSV but there is no one size fits all solution when it comes to data analysis.

You will have to convert data to a format that you can use.
With excel or google spreadsheet, you can use text splitting to go from lap time to milliseconds, for example.

There is some coding skills to acquire here.
Sorry that I don't have the solution here, if I have time to look into this data, I will provide a Google spreadsheet template.

Share what kind of data analysis you would like to do, I will try to add these logics to the spreadsheet.
md-quotelink
medal 5000
4 years 354 days ago
I'm a finance analyst, so thought I'd love delving into the numbers on this game.

As soon as i started, i threw together a program what would take in the lap times on practise (would do 2 at one tyre compound, 3 at another, then average out the times (all on perfect setup since it's pretty much fixed + a driver offset), factor in the pit duration for a given track, tyre wear per lap etc to give the optimal number of stints and on which tyres etc.

My findings... There's so many hidden variables on this game that there's no point for a lot of it. At least not until you hit lvl 18+ where your car is going to be around a consistent level each time you race at a certain track. The program i made was 100% useless because it turns out that practice times might as well be randomly generated. You'll have two identical drivers, both setting same times in practise (within around 0.1s). They both set nearly exactly 0.5s better lap times on softs vs mediums... You start one on mediums, one on soft, same fuel, then end up with the guy on mediums setting the same lap times as the guy on softs. 

Add to that the drivers being wildy inconsistent at lower levels means you have too many unknowns and too small of a data set to actually make any use of that data available to download. You'd need 20+ races at the same track with the same car setup, with similar weather to begin to draw any conclusions on what hidden variables are in play and if they're random, or based on a factor somewhere, which is only available to high level players who have stable crew/car setups each season

I'm sure when if i stick it out long enough to hit a high level I'll have some fun with the data but for now, i don't bother.
md-quotelink
medal 5000
4 years 354 days ago
@Wambou Wambou I tried this solution you write about - change text to millisecounds - but i thought that there is a simpler way. I would like to get a window with level of fuel and tires where is the most optimal situation for good time lap. 

@jack pi Thank you for your experienced point of view. There is a lot of truth and I didn't think about it in this way before. But I think that we could collect data from this moment and we will be able to compare all of data in the future. By the way it is really interesting so I want to do it for my own satisfaction :) but i need help with this. 
md-quotelink
medal 5000
4 years 354 days ago
Go for it if you find it interesting for sure!

The reason I don't collect for now is because between this race on the current track and the next one next season I have all the following variables that will affect the race time and stop me from drawing any conclusions: 

- Was it down to my change in tactic?
- Was it down to my drivers being slightly more trained than last season?
- Was it down to my car being slightly better this season?
- Was it down to the slight change in weather?
- Was it down to my guy being stuck in traffic for a few seconds less this race?
- Was it down to my driver hitting an extra DRS this race?
- was it just simply a random variable build in somewhere?
- Was it a combination of all the above? If so, what combination etc

I understand them giving us this info, but without realistic practice times, which is the only data we have to go on in early leagues, its next to useless for improving your strategy :(
md-quotelink
medal 5000
4 years 350 days ago
use google chrome download the csv file it will download into your downloads section
in excel click on the data tab
then click on import from text
go to your downloads section and import the csv file 

click on the commas option and press finish

make sure the cells are which you want the data to go in are correct and your done
md-quotelink
medal 5000
4 years 350 days ago
@Andy Higgins 
That's clear, the problem is with interpretation of data. 
md-quotelink
medal 5000
4 years 350 days ago
What you want to get from it
md-quotelink
medal 5000
4 years 323 days ago
Could the following be added to the downloadable CSV?

Per lap:-


  • Tyre Temp - as +/- degrees from optimum

  • The average push level for the lap i.e.




          if 1 = No Push 5 = Push Hard
pushing at level 2 for half the lap but level 1 for the other half would result in a value of 1.5 for that lap.


  • Time in seconds to car in front, at the moment it's just to the leader. This can help diagnose if the car is slower on a particular lap because of tyres going off, or being held up by a slow car on e.g. Monaco



For Whole Race/Day (Maybe on a different Tab):-


  • The setup for each practice lap and the time achieved, tyre wear, fuel usage.

  • The setup used for the race.

  • Qualifying lap, or position and time.




md-quotelink
angle-double-left ios-arrow-back 1 ios-arrow-forward angle-double-right

You must be logged in to post a reply.