Git Product home page Git Product logo

tadw's Issues

node embeddings are all zero

Hi, Thanks for the python implementation of TADW. I try to evaluate the result of TADW on giraffe dataset. I make the following changes to the codes:

  • chage the file path of --edge-path, --feature-path,--output-path to giraffe dataset.
  • change --features to dense.

And all other parameters keeps unchanged.

Here are embeddings I get:

id,X_0,X_1,X_2,X_3,X_4,X_5,X_6,X_7,X_8,X_9,X_10,X_11,X_12,X_13,X_14,X_15,X_16,X_17,X_18,X_19,X_20,X_21,X_22,X_23,X_24,X_25,X_26,X_27,X_28,X_29,X_30,X_31,X_32,X_33,X_34,X_35,X_36,X_37,X_38,X_39,X_40,X_41,X_42,X_43,X_44,X_45,X_46,X_47,X_48,X_49,X_50,X_51,X_52,X_53,X_54,X_55,X_56,X_57,X_58,X_59,X_60,X_61,X_62,X_63
0.0,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,17.5561640193,19.0393369854,17.9668565593,17.8095968136,15.7288687625,18.2842039046,18.8672905625,19.0908401457,17.3428634989,17.9861489809,19.2221152214,18.8880063305,18.9347242007,16.3956388766,18.0626110202,17.7801313876,19.6118153724,19.6522067374,15.5831001956,17.9466391123,18.9412024245,18.6910995146,18.4319755594,17.7823144487,16.1617267138,16.632035156,17.7868636472,19.144048575,16.2448227439,15.8837429851,16.8948444098,20.7434825928
1.0,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,19.418857302,21.1776514057,16.7330740863,20.7961903049,17.4019876294,21.4364204382,20.5935195948,20.8804407443,19.4307565643,21.2201610956,19.8779685732,21.658861326,20.3617437599,18.0326615246,19.5013417891,20.4816495277,20.1702035628,19.1156188389,17.2827884159,20.5683601985,22.6392520365,20.0903648231,19.9298201164,21.3580573765,18.7907836501,19.7752476486,19.9652307795,20.5689666432,18.7515709792,17.0995186355,18.5041452107,20.1793805727
2.0,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,15.8655501853,15.2614582941,15.1237360692,14.5464554457,15.518018525,15.1216947038,14.9469130927,15.2269697516,14.2152368637,15.2137422453,16.0391038165,15.1560322943,16.9188702633,15.134795549,14.7140644724,14.9903461376,16.4724884343,15.785330147,14.6277017312,16.4522976424,16.0744194761,15.5419878179,15.7469767053,14.1617998808,14.5053627042,14.7402181088,15.7792771715,14.8755418723,15.1000829619,13.7796906747,15.21955654,16.8494631693
3.0,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,25.2035641617,25.7037428377,24.1046696175,23.1220193021,23.7414492738,23.8769181891,23.2346393145,24.6325339183,23.9417420937,25.4818110078,26.0248145741,26.4453890377,25.7432762261,23.3298515663,24.8734525125,24.0201878555,24.1759821777,25.4479371142,23.0832849934,25.8504544394,25.8230451379,26.0038499866,25.2155392266,24.2239254842,24.2414188799,24.0078842097,25.1042005095,24.3209910274,24.6705904816,21.8627201866,24.0299475081,24.0038341944
4.0,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,1e-15,10.4560723562,12.5196933928,10.6806982438,13.076437942,9.72079032744,8.94961917529,12.7237842975,10.9228435878,10.3742308348,12.0424505048,12.2636346198,12.6850735313,9.37093654152,7.06472852324,13.0849199387,12.7714280378,11.8251953826,10.8370930474,10.0865815703,11.7727932304,11.2289073837,10.3189405835,12.5184795748,14.4196606615,11.7215457371,10.2559107664,11.5965745013,12.1475858113,9.22081082777,11.1381676022,12.4477135617,12.3033616545
...

It seems that the first half part dimension embeddings are nearly zero. Can you solve this problem for me? Thanks.

Good work! But one question.

Hi @benedekrozemberczki,

Why I cannot train similar nodes to similar embeddings?
So what I did is training the node embedding twice on the same dataset. Same node should output similar embeddings right? I experimented with the given samples, not work. Then I created a even simpler dataset crafted by myself, don't work neither.

Leo

Where is Text-feature matrix? what is json file?

TADW/src/main.py

Line 14 in e7ec745

X = read_features(args.feature_path) 

def read_features(feature_path):
    """
    Method to get dense node feaures.
    :param feature_path: Path to the node features.
    :return features: Node features.
    """
    features = pd.read_csv(feature_path)
    features = np.array(features)[:, 1:].transpose()
    return features

I think this should be the text tfidf information reduction module . Accorind to the comment it is for getting dense node features. According to the readme, the json file is the feature matrix. But it is the feature matrix of what? nodes feature matrix or text feature matrix? if it is the network information, where is the text feature matrix? if it is a text feature matrix, why it is not tf-idf of the text-data. How is the json file come? Could you show the code for producing json file?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.