centipawn_loss_calculator's Introduction
centipawn_loss_calculator's People
Forkers
rubenbelem sergioleao gabrieldefreire jmolinahn asbenicio dkappe jmararr1 tombrau ldsalomone henriklimseth rdparker athospd blankcanvasstudio zenthwolf maxmatkovski raphaelamorim luke-lite masbrutus kidkal edoson gabrielpiazzalunga ecyl-hub tomplt ekinyua cardosojrcentipawn_loss_calculator's Issues
could you clarify Evaluation calculations in the csv?
Hi,
trying to build upon your work a little bit. Can you help me understand how exactly are evaluations calculated? Looking into raw csv data, I noticed that first eval value is always positive, but varies:
df_cpwn['evals']=df_cpwn['Evaluations List'].apply(lambda s: list(map(int, s.strip('[]').split(','))))
df_cpwn.evals.apply(lambda x: x[0]).describe()
count 7568.000000
mean 39.310386
std 6.811494
min 16.000000
25% 35.000000
50% 39.000000
75% 44.000000
max 87.000000
Name: evals, dtype: float64
So first eval cannot be move 1, but it also cannot be something like move 10, since it's hard to believe that in 7k games black not even once had any advantage on that move.
edit: TIL, Stockfish evaluates the initial position as something random between 16 and 87.
Inconsistent Formatting of .csv files
Not all games in 15982_games_with_centipawn_metrics.csv contain the analysis of the positions. I can write some code to prevent these games from being inserted into the csv, which would prevent the inconsistent formatting. But I want to make sure their inclusion is not intentional before I try to pull the update.
Add script for plots from dataframe
Create plots using tools such as seaborn
to automatically generate the data plot of players
I suggest a dispersion plot of Cp loss vs. Rating for each player, calculating the tendency line and the R^2 to check the correlation. It's clear from the results that Niemann has little or no correlation between Cp loss and rating.
Seaborn has all the tools to do all that automatically and generate beautiful plots.
MPV feature used ?
after tinkering your script a bit, i have expected results and it works with several engines .. i suppose all (proper) UCI engines can be used, also those which can do Multi-PV, but are such PV lines included in the results? I mean, many positions can have many 'good' moves with almost same eval ..
0-CPL moves
Great analysis. I just had a quick look at average number of 0-CPL moves per game, and Niemann is coming out higher than the other players here. The 0-CPL moves are going to be more indicative of cheating.
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.