Youtube friendships

This is the friendship network of the video-sharing site Youtube. Nodes are users and an undirected edge between two nodes indicates a friendship.


Internal namecom-youtube
NameYoutube friendships
Data source
AvailabilityDataset is available for download
Consistency checkDataset passed all tests
Online social network
Node meaningUser
Edge meaningFriendship
Network formatUnipartite, undirected
Edge typeUnweighted, no multiple edges
LoopsDoes not contain loops


Size n =1,134,890
Volume m =2,987,624
Loop count l =0
Wedge count s =1,474,482,560
Claw count z =5,743,242,041,025
Cross count x =32,746,672,524,915,828
Triangle count t =3,056,386
Square count q =468,774,021
4-Tour count T4 =9,654,097,656
Maximum degree dmax =28,754
Average degree d =5.265 05
Fill p =4.639 26 × 10−6
Size of LCC N =1,134,890
Diameter δ =24
50-Percentile effective diameter δ0.5 =4.860 22
90-Percentile effective diameter δ0.9 =6.934 78
Median distance δM =5
Mean distance δm =5.548 06
Gini coefficient G =0.714 126
Balanced inequality ratio P =0.213 532
Relative edge distribution entropy Her =0.877 112
Power law exponent γ =2.429 00
Tail power law exponent γt =2.141 00
Degree assortativity ρ =−0.036 909 9
Degree assortativity p-value pρ =0.000 00
Clustering coefficient c =0.006 218 56
Spectral norm α =210.395
Non-bipartivity bA =0.161 013
Normalized non-bipartivity bN =0.002 045 96
Algebraic non-bipartivity χ =0.004 266 35
Spectral bipartite frustration bK =0.000 202 579


Degree distribution

Cumulative degree distribution

Lorenz curve

Spectral distribution of the adjacency matrix

Spectral distribution of the normalized adjacency matrix

Spectral distribution of the Laplacian

Spectral graph drawing based on the adjacency matrix

Spectral graph drawing based on the Laplacian

Spectral graph drawing based on the normalized adjacency matrix

Hop distribution

Clustering coefficient distribution


Matrix decompositions plots



[1] Jérôme Kunegis. KONECT – The Koblenz Network Collection. In Proc. Int. Conf. on World Wide Web Companion, pages 1343–1350, 2013. [ http ]
[2] Jaewon Yang and Jure Leskovec. Defining and evaluating network communities based on ground-truth. In Proc. ACM SIGKDD Workshop on Min. Data Semant., page 3, 2012.