Graphical Modelling and Causal Inference
Published April 2016 | Version v6
Dataset Open

Cause-Effect Pairs from MPI Tübingen

Description

This collection of cause-effect pair datasets was created by the Max-Planck-Institute in Tuebingen, Germany (https://webdav.tuebingen.mpg.de/cause-effect/).

 

Task: Identify for each pair the causal direction using the observed samples only.

 

Summary: 

  • Size of collection: 108 datasets with 2 columns each of various sizes
  • Task: Causal Discovery Problem
  • Data Type: Mixed Data
  • Dataset Scope: Collection of Datasets
  • Ground Truth: Known Graph
  • Temporal Structure: Static Data
  • License: CC BY-NC 4.0 (for exceptions see below)
  • Missing Values: No Missing Values

 

Missingness Statement: There are no missing values.

 

Collection: The database contains the following datasets.

Nr. Variable 1 Variable 2 Origin of datasets Direction
1 Altitude Temperature DWD ->
2 Altitude Precipitation DWD ->
3 Longitude Temperature DWD ->
4 Altitude Sunshine hours DWD ->
5 Age Length Abalone ->
6 Age Shell weight Abalone ->
7 Age Diameter Abalone ->
8 Age Height Abalone ->
9 Age Whole weight Abalone ->
10 Age Shucked weight Abalone ->
11 Age Viscera weight Abalone ->
12 Age Wage per hour census income ->
13 Displacement Fuel consumption auto-mpg ->
14 Horse power Fuel consumption auto-mpg ->
15 Weight Fuel consumption auto-mpg ->
16 Horsepower Acceleration auto-mpg ->
17 Age Dividends from stocks census income ->
18 Age Concentration GAG GAGurine (from R package MASS) ->
19 Current duration Next interval geyser ->
20 Latitude Temperature DWD ->
21 Longitude Precipitation DWD ->
22 Age Height arrhythmia ->
23 Age Weight arrhythmia ->
24 Age Heart rate arrhythmia ->
25 Cement Compressive strength concrete_data ->
26 Blast furnace slag Compressive strength concrete_data ->
27 Fly ash Compressive strength concrete_data ->
28 Water Compressive strength concrete_data ->
29 Superplasticizer Compressive strength concrete_data ->
30 Coarse aggregate Compressive strength concrete_data ->
31 Fine aggregate Compressive strength concrete_data ->
32 Age Compressive strength concrete_data ->
33 Alcohol consumption Mean corpuscular volume  liver disorders ->
34 Alcohol consumption Alkaline phosphotase  liver disorders ->
35 Alcohol consumption Alanine aminotransferase liver disorders ->
36 Alcohol consumption Aspartate aminotransferase liver disorders ->
37 Alcohol consumption Gamma-glutamyl transpeptdase liver disorders ->
38 Age Body mass index pima indian diabetes ->
39 Age Serum insulin pima indian diabetes ->
40 Age Diastolic blood pressure pima indian diabetes ->
41 Age Plasma glucose concentration pima indian diabetes ->
42 Day of the year Temperature D. Janzing ->
43 Temperature at t Temperature at t+1 ncep-ncar ->
44 Pressure at t Pressure at t+1 ncep-ncar ->
45 Sea level pressure at t Sea level pressure at t+1 ncep-ncar ->
46 Relative humidity at t Relative humidity at t+1 ncep-ncar ->
47 Number of cars Type of day traffic <-
48 Indoor temperature Outdoor temperature Hipel & Mcleod <-
49 Ozone concentration Temperature Bafu <-
50 Ozone concentration Temperature Bafu <-
51 Ozone concentration Temperature Bafu <-
52 (Temp, Press, SLP, Rh) (Temp, Press, Slp, Rh) ncep-ncar <-
53 Ozone concentration (Wind speed, Radiation, Temperature) environmental <-
54 (Displacement, Horsepower, Weight)   (Fuel consumption, Acceleration) auto-mpg ->
55 Ozone concentration (16-dim.) Radiation (16-dim.) Bafu <-
56 Female life expectancy, 2000-2005 Latitude UNdata <-
57 Female life expectancy, 1995-2000 Latitude UNdata <-
58 Female life expectancy, 1990-1995 Latitude UNdata <-
59 Female life expectancy, 1985-1990 Latitude UNdata <-
60 Male life expectancy, 2000-2005 Latitude UNdata <-
61 Male life expectancy, 1995-2000 Latitude UNdata <-
62 Male life expectancy, 1990-1995 Latitude UNdata <-
63 Male life expectancy, 1985-1990 Latitude UNdata <-
64 Drinking water access Infant mortality UNdata ->
65 Stock return of Hang Seng Bank Stock return of HSBC Hldgs Yahoo database ->
66 Stock return of Hutchison Stock return of Cheung kong Yahoo database ->
67 Stock return of Cheung kong  Stock return of Sun Hung Kai Prop. Yahoo database ->
68 Bytes sent  Open http connections  P. Stark & Janzing <-
69 Inside temperature Outside temperature J.M. Mooij <-
70 Parameter Answer Armann & Buelthoff ->
71 Symptoms (6-dim.) Classification of disease (2-dim.) Acute Inflammations  ->
72 Sunspots Global mean temperature sunspot data ->
73 CO2 emissions Energy use UNdata <-
74 GNI per capita Life expectancy UNdata ->
75 Under-5 mortality rate GNI per capita UNdata <-
76 Population growth Food consumption growth Food and Agriculture Organization of the United Nations     ->     
77 Temperature Solar radiation B. Janzing        <-
78 PPFD Net Ecosystem Productivity Moffat A.M. ->
79 Net Ecosystem Productivity Diffuse PPFDdif Moffat A.M. <-
80 Net Ecosystem Productivity Direct PPFDdir Moffat A.M. <-
81 Temperature Local CO2 flux, BE-Bra Mahecha, M. ->
82 Temperature Local CO2 flux, DE-Har Mahecha, M. ->
83 Temperature Local CO2 flux, US-PFa Mahecha, M. ->
84 Employment Population http://www.spatial-econometrics.com <-
85 Time of measurement Protein content of milk http://www.maths.lancs.ac.uk/Software/Oswald/ ->
86 Size of apartment Monthly rent J.M. Mooij ->
87 Temperature Total snow http://www.mldata.org/repository/data/viewslug/whistler-daily-snowfall/ ->
88 Age Relative spinal bone mineral density bone dataset of R ElemStatLearn package ->
89 root decomposition Oct (grassl) root decomposition Oct (grassl) Solly et al (2014). Plant and Soil, 382(1-2), 203-218. <-
90 root decomposition Oct (forest) root decomposition Oct (forest) Solly et al (2014). Plant and Soil, 382(1-2), 203-218. <-
91 clay cont. in soil (forest) soil moisture Solly et al (2014). Plant and Soil, 382(1-2), 203-218. ->
92 organic carbon in soil (forest) clay cont. in soil (forest) Solly et al (2014). Plant and Soil, 382(1-2), 203-218. <-
93 precipitation runoff MOPEX (ftp://hydrology.nws.noaa.gov/pub/gcip/mopex/US_Data/Us_438_Daily/) ->
94 hour of day temperature S. Armagan Tarim ->
95 hour of day electricity load S. Armagan Tarim ->
96 temperature electricity load S. Armagan Tarim ->
97 speed at the beginning speed at the end D. Janzing ->     
98 speed at the beginning speed at the end D. Janzing ->    
99 language test score social-economic status family nlschools dataset of R MASS package <-
100 cycle time of CPU performance cpus dataset of R MASS package ->
101 grey value of a pixel brightness of the screen D. Janzing ->      
102 position of a ball time for passing a track segment D. Janzing ->
103 position of a ball time for passing a track segment D. Janzing ->
104 time for passing 1. segment time for passing 2. segment D. Janzing ->
105 pixel vector of a patch total brightness at the screen D. Janzing -> 
106 time required for one round voltage D. Janzing <-   
107 strength of contrast answer correct or not Schuett, edited by D. Janzing ->    
108 time for 1/6 rotation temperature D. Janzing <-    

 

Files:

  • pairs.zip: collection of cause-effect pair datasets
  • ground_truth.txt: ground truth of causal direction

License:

  • The datasets by D. Janzing stand by CC-BY 4.0
  • For the remaining datasets, please check individually

Files

ground_truth.txt

Files (1.0 MB)

Name Size Download all
md5:b31280745498f5938c3872da27835d8b
1.0 kB Preview Download
md5:29cd8d0a49f80f9423ee7a5fe7c70df1
1.0 MB Preview Download

Additional details

Identifiers

URL
https://webdav.tuebingen.mpg.de/cause-effect/

References

  • J. M. Mooij, J. Peters, D. Janzing, J. Zscheischler, B. Schoelkopf: "Distinguishing cause from effect using observational data: methods and benchmarks", Journal of Machine Learning Research 17(32):1-102, 2016