Comments (2)
It's possible that there's a bug somewhere, but I can't obtain good results by running this package.
I implemented a simple script:
using CSV
using DataFrames
# using YAML
using TransferEntropy
samples_filename = ARGS[1]
seed = parse(Int, ARGS[2])
dim_x = parse(Int, ARGS[3])
dim_y = parse(Int, ARGS[4])
# metadata = YAML.load_file(metadata_filename)
# show(metadata)
# dim_x = metadata["dim_x"]
# dim_y = metadata["dim_y"]
df = CSV.File(samples_filename) |> DataFrame
df = df[df.seed .== seed, :]
x_samples = Dataset(Matrix(df[:, 2:1+dim_x]))
y_samples = Dataset(Matrix(df[:, 2+dim_x: 1+dim_x+dim_y]))
# show(x_samples)
# show(y_samples)
mi = mutualinfo(x_samples, y_samples, Kraskov())
println(mi)
mi = mutualinfo(x_samples, y_samples, Kraskov1())
println(mi)
mi = mutualinfo(x_samples, y_samples, Kraskov2())
println(mi)
mi = mutualinfo(x_samples, y_samples, KozachenkoLeonenko())
println(mi)
est = VisitationFrequency(RectangularBinning(0.1))
mi = mutualinfo(x_samples, y_samples, est)
println(mi)
est = VisitationFrequency(RectangularBinning(0.2))
mi = mutualinfo(x_samples, y_samples, est)
println(mi)
est = VisitationFrequency(RectangularBinning(0.5))
mi = mutualinfo(x_samples, y_samples, est)
println(mi)
est = VisitationFrequency(RectangularBinning(1))
mi = mutualinfo(x_samples, y_samples, est)
println(mi)
and generated a task using split multinormal sampler:
dim_x: 1
dim_y: 2
mi_true: 0.5108256237659909
n_samples: 5000
task_id: some-correlation
Then I ran the script:
$ julia mi_estimator.jl testdir/samples.csv 1 1 2
and obtained:
-Inf
2.0348837647400924
2.405459973695475
-Inf
3.5470532390132483
1.7091615702982228
0.8061644481070918
0.0
Some ideas:
- Perhaps the unit is different (and it's not nats, but something else).
- Maybe the conversion from the data frame to data set is wrong.
from bmi.
This has been added using TransferEntropy.jl
library.
from bmi.
Related Issues (20)
- Principled approach to handling NaNs HOT 1
- Benchmark versioning
- Improve the documentation
- Add difference of cross-entropies estimator
- Add the tasks based on fine distributions
- Add dropout to neural estimators
- Inconsistent `.mutual_information` method HOT 1
- Rename fine distributions to BMMs
- Allow for Python 3.11 and 3.12
- Changes to the manuscript HOT 2
- Add MI estimators in R HOT 2
- Wrap the estimators to make them easier to run HOT 1
- Change the package API HOT 1
- Better tests for estimators
- Implement early stopping for neural estimators HOT 1
- Update github readme HOT 1
- Clean up imports HOT 1
- Minibenchmarks HOT 1
- Error raised during smoothing if training is too short
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 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. 📊📈🎉
-
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 ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from bmi.