{'diabetes_readmission_causal_group_dro'}
{'validation': 0.5762949136724218, 'id_test': 0.5838581758805692, 'ood_test': 0.5061999686077538, 'ood_validation': 0.4929378531073446, 'train': 0.5752158189454036}
Train Datashifts with Representations:{'informativeness': 0.02134312385762539, 'invariance': 0.3507385551022407, 'label_shift': 0.003375592508434266, '(latent)_covariate_shift': 0.3735572846293687, 'concept_shift': 0.23972031074062852, 'residual': 0.031100222626243262, 'overall': -0.11652009312115101}
Test Datashifts with Representations:{'informativeness': 0.017858193987684266, 'invariance': 0.0008342452149945886, 'label_shift': 0.003414277603891494, '(latent)_covariate_shift': 0.0008315541286126125, 'concept_shift': 0.06011014291591765, 'residual': 0.021013148949305113, 'overall': -0.031504233160824915}
{'diabetes_readmission_arguablycausal_group_dro'}
{'validation': 0.5909939337377508, 'id_test': 0.5978539771401913, 'ood_test': 0.5349631141108147, 'ood_validation': 0.5291313559322034, 'train': 0.5921896873541764}
Train Datashifts with Representations:{'informativeness': 0.09110577418696982, 'invariance': 0.3868877762342364, 'label_shift': 0.003375592508434266, '(latent)_covariate_shift': 0.42805342684664616, 'concept_shift': 0.010234730698678526, 'residual': 0.08777079348523491, 'overall': 0.020488236912817648}
Test Datashifts with Representations:{'informativeness': 0.08336788964757917, 'invariance': 0.0001442695040896813, 'label_shift': 0.003414277603891494, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.0019115709291783868, 'residual': 0.07922499258802362, 'overall': 0.004822115644867259}
{'diabetes_readmission_group_dro'}
{'validation': 0.6105926271581894, 'id_test': 0.609050618147889, 'ood_test': 0.5528763145503061, 'ood_validation': 0.5483757062146892, 'train': 0.6089885674288381}
Train Datashifts with Representations:{'informativeness': 0.07591370474845006, 'invariance': 0.6924163193568724, 'label_shift': 0.003375592508434266, '(latent)_covariate_shift': 0.7391739365964914, 'concept_shift': 0.0029432995949840437, 'residual': 0.07598938562010517, 'overall': 0.02351927420487948}
Test Datashifts with Representations:{'informativeness': 0.08530585287071288, 'invariance': 7.213475204540125e-05, 'label_shift': 0.003414277603891494, '(latent)_covariate_shift': 0.0, 'concept_shift': 4.805139755722377e-16, 'residual': 0.07863973158858281, 'overall': 0.008337192708052873}
{'acsfoodstamps_causal_group_dro'}
{'validation': 0.8183448433744134, 'id_test': 0.8156636312763901, 'ood_test': 0.7819059699660379, 'ood_validation': 0.7902780335113239, 'train': 0.8159655208594985}
Train Datashifts with Representations:{'informativeness': 0.07612903230936671, 'invariance': 2.1300276683229673, 'label_shift': -0.0058060624703405814, '(latent)_covariate_shift': 2.2551408215399418, 'concept_shift': 0.1109291730926157, 'residual': 0.07742204904863265, 'overall': 0.0028959420877431585}
Test Datashifts with Representations:{'informativeness': 0.0791482334937432, 'invariance': -0.0001441401086572871, 'label_shift': -0.0015958378682764684, '(latent)_covariate_shift': 7.272725755671429e-05, 'concept_shift': 0.13819558371021654, 'residual': 0.07580521560013197, 'overall': -0.06644425921252826}
{'acsfoodstamps_arguablycausal_group_dro'}
{'validation': 0.8351584061454717, 'id_test': 0.831217886757898, 'ood_test': 0.8038585866852163, 'ood_validation': 0.8081384643712024, 'train': 0.832322763418535}
Train Datashifts with Representations:{'informativeness': 0.16610914820375358, 'invariance': 1.032837034381398, 'label_shift': -0.0058060624703405814, '(latent)_covariate_shift': 1.0737180191540359, 'concept_shift': -0.007852858081954465, 'residual': 0.10288773005606973, 'overall': 0.08468530833980974}
Test Datashifts with Representations:{'informativeness': 0.17358127908031432, 'invariance': 0.004219882994600474, 'label_shift': -0.0015958378682764684, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.003642804978244632, 'residual': 0.11126896835435027, 'overall': 0.05758304780540324}
{'acsfoodstamps_group_dro'}
{'validation': 0.8453584646495479, 'id_test': 0.8432365060792593, 'ood_test': 0.8156430295838618, 'ood_validation': 0.8147670778862088, 'train': 0.8442556492818966}
Train Datashifts with Representations:{'informativeness': 0.20557628648578768, 'invariance': 0.9846413396056644, 'label_shift': -0.0058060624703405814, '(latent)_covariate_shift': 1.020286667829203, 'concept_shift': -0.00533134329753265, 'residual': 0.11261607501464756, 'overall': 0.11054551599650554}
Test Datashifts with Representations:{'informativeness': 0.21233539923946493, 'invariance': 0.003570670226199391, 'label_shift': -0.0015958378682764684, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.001875503553155606, 'residual': 0.13032090729081133, 'overall': 0.07849348612483784}
{'brfss_diabetes_causal_group_dro'}
{'validation': 0.8727420826765417, 'id_test': 0.8733539482615582, 'ood_test': 0.8256898387489898, 'ood_validation': 0.8263613623244179, 'train': 0.8726770320313149}
Train Datashifts with Representations:{'informativeness': 0.02512011531236209, 'invariance': 0.0067136935442241395, 'label_shift': 0.017554929267790703, '(latent)_covariate_shift': 0.00573568932638694, 'concept_shift': 0.041116542311724756, 'residual': 0.0198589348242304, 'overall': -0.00700862814275394}
Test Datashifts with Representations:{'informativeness': 0.038174019881482064, 'invariance': 0.14216570782451035, 'label_shift': 0.016442664018816142, '(latent)_covariate_shift': 0.15030095725474696, 'concept_shift': 0.1194095969567411, 'residual': 0.033064997143921575, 'overall': -0.042306819016283685}
{'brfss_diabetes_arguablycausal_group_dro'}
{'validation': 0.8745233339386236, 'id_test': 0.8727734948908001, 'ood_test': 0.8258053731343283, 'ood_validation': 0.828748280605227, 'train': 0.8752812802753529}
Train Datashifts with Representations:{'informativeness': 0.08877230313916598, 'invariance': 0.0008656170245350446, 'label_shift': 0.004410807874824522, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.0007934822724894832, 'residual': 0.08877205139301728, 'overall': 0.001376106035048702}
Test Datashifts with Representations:{'informativeness': 0.08529543964610285, 'invariance': 0.1658115482671657, 'label_shift': 0.007387840843629615, '(latent)_covariate_shift': 0.17286341645454956, 'concept_shift': 0.005454220959007892, 'residual': 0.09111661062627814, 'overall': -0.0013284269441724977}
{'brfss_diabetes_group_dro'}
{'validation': 0.8737639698235303, 'id_test': 0.87121349687175, 'ood_test': 0.830380895522388, 'ood_validation': 0.828877235213205, 'train': 0.8744104850350124}
Train Datashifts with Representations:{'informativeness': 0.13109220433679114, 'invariance': 0.0003606737602225215, 'label_shift': 0.004410807874824522, '(latent)_covariate_shift': 0.0, 'concept_shift': 3.6067376022940885e-05, 'residual': 0.116800670971895, 'overall': 0.016298566734185643}
Test Datashifts with Representations:{'informativeness': 0.1376574287146991, 'invariance': 0.2317221068380068, 'label_shift': 0.007387840843629615, '(latent)_covariate_shift': 0.24677524767710407, 'concept_shift': 0.0017708909367111646, 'residual': 0.13269550741322492, 'overall': 0.015296966674482049}
{'acsincome_causal_group_dro'}
{'validation': 0.7050153466443059, 'id_test': 0.7066309740785743, 'ood_test': 0.6690071267668717, 'ood_validation': 0.6667457024303497, 'train': 0.7044947366672388}
Train Datashifts with Representations:{'informativeness': 0.12633377716460964, 'invariance': 1.3295548290371346, 'label_shift': 0.015213404934141922, '(latent)_covariate_shift': 1.3256671440187142, 'concept_shift': 0.2081000517395108, 'residual': 0.12486489912969634, 'overall': -0.09691828787698129}
Test Datashifts with Representations:{'informativeness': 0.1579812789045621, 'invariance': -0.00015923018689335285, 'label_shift': 0.003482233702930959, '(latent)_covariate_shift': -0.00017586876160316614, 'concept_shift': 0.1020488393141629, 'residual': 0.15665857161774005, 'overall': -0.04796891480614887}
{'acsincome_arguablycausal_group_dro'}
{'validation': 0.820206942378888, 'id_test': 0.8197777440259214, 'ood_test': 0.79498359921487, 'ood_validation': 0.7970361588618851, 'train': 0.8215482195957198}
Train Datashifts with Representations:{'informativeness': 0.2966450985134879, 'invariance': 1.4324576578991466, 'label_shift': 0.015213404934141922, '(latent)_covariate_shift': 1.5019226064291717, 'concept_shift': -0.0082087542753961, 'residual': 0.1209858016232189, 'overall': 0.22210285076005062}
Test Datashifts with Representations:{'informativeness': 0.41247940097683056, 'invariance': 0.011546257497527109, 'label_shift': 0.003482233702930959, '(latent)_covariate_shift': 5.773702524787593e-06, 'concept_shift': 0.016832644389572056, 'residual': 0.20947884275112616, 'overall': 0.19055511098488265}
{'acsincome_group_dro'}
{'validation': 0.8235167547384742, 'id_test': 0.8231571486431754, 'ood_test': 0.814440594907194, 'ood_validation': 0.811262596324837, 'train': 0.8248081871779882}
Train Datashifts with Representations:{'informativeness': 0.25105606786879797, 'invariance': 0.9871663627667863, 'label_shift': 0.015213404934141922, '(latent)_covariate_shift': 1.026794920136384, 'concept_shift': -0.008238033350985895, 'residual': 0.08046665057518258, 'overall': 0.20212941512097815}
Test Datashifts with Representations:{'informativeness': 0.3634298338006934, 'invariance': 0.020378067452557647, 'label_shift': 0.003482233702930959, '(latent)_covariate_shift': 2.633591064131417e-06, 'concept_shift': 0.009521787269867552, 'residual': 0.137755113459377, 'overall': 0.21246722662710135}
{'acsunemployment_causal_group_dro'}
{'validation': 0.9665910612032423, 'id_test': 0.966176060483996, 'ood_test': 0.9484875711290806, 'ood_validation': 0.9493949394939494, 'train': 0.9659411858574622}
Train Datashifts with Representations:{'informativeness': 0.008889289888310349, 'invariance': 0.17330484353519904, 'label_shift': 0.007477877913284072, '(latent)_covariate_shift': 0.17401830377010916, 'concept_shift': 0.13289598583046222, 'residual': 0.007571576516442423, 'overall': -0.06103461046926608}
Test Datashifts with Representations:{'informativeness': 0.012869068203874193, 'invariance': 0.33560240192362895, 'label_shift': -0.011390759908765893, '(latent)_covariate_shift': 0.3412145324477917, 'concept_shift': 0.12281004397985452, 'residual': 0.009936912095900247, 'overall': -0.061362180574254854}
{'acsunemployment_arguablycausal_group_dro'}
{'validation': 0.9722614709600654, 'id_test': 0.9717039010937936, 'ood_test': 0.960375524873022, 'ood_validation': 0.961001100110011, 'train': 0.971378418701788}
Train Datashifts with Representations:{'informativeness': 0.038226652133188405, 'invariance': 0.3486196292032117, 'label_shift': 0.007477877913284072, '(latent)_covariate_shift': 0.3508445941017723, 'concept_shift': 0.0055670627613812525, 'residual': 0.014000228566774056, 'overall': 0.026294313591646072}
Test Datashifts with Representations:{'informativeness': 0.08438659103965089, 'invariance': 0.340895846167844, 'label_shift': -0.011390759908765893, '(latent)_covariate_shift': 0.34199917821767584, 'concept_shift': 0.022405816478042743, 'residual': 0.04123578096919368, 'overall': 0.0268041879019688}
{'acsunemployment_group_dro'}
{'validation': 0.9728811878733794, 'id_test': 0.9723545998202832, 'ood_test': 0.9610539633643215, 'ood_validation': 0.9611661166116612, 'train': 0.9723932035751414}
Train Datashifts with Representations:{'informativeness': 0.10630972958168233, 'invariance': 0.2973841676011296, 'label_shift': 0.007477877913284072, '(latent)_covariate_shift': 0.29147486470790385, 'concept_shift': 0.0005772073505795653, 'residual': 0.08403898099547724, 'overall': 0.022766432420944463}
Test Datashifts with Representations:{'informativeness': 0.17465781455062043, 'invariance': 0.38587005088830467, 'label_shift': -0.011390759908765893, '(latent)_covariate_shift': 0.37848464474723886, 'concept_shift': 0.01460999630828402, 'residual': 0.11921661666125584, 'overall': 0.038748116710306715}
{'assistments_causal_group_dro'}
{'validation': 0.9236886924814117, 'id_test': 0.9249191569817606, 'ood_test': 0.45120671563483733, 'ood_validation': 0.5518867924528302, 'train': 0.9237861578241016}
Train Datashifts with Representations:{'informativeness': 0.3232800851692643, 'invariance': 1.820916586557217, 'label_shift': 0.02660028896165906, '(latent)_covariate_shift': 1.8623595840603233, 'concept_shift': 0.17200585609038385, 'residual': 0.09903799677958479, 'overall': 0.17226080357687024}
Test Datashifts with Representations:{'informativeness': 0.21482956376893458, 'invariance': 2.007761040649984, 'label_shift': 0.02329921323212632, '(latent)_covariate_shift': 2.065693256800828, 'concept_shift': 0.0019402717203180037, 'residual': 0.24064181743567517, 'overall': 0.013833325164585603}
{'assistments_arguablycausal_group_dro'}
{'validation': 0.9211414809090431, 'id_test': 0.9221281033590181, 'ood_test': 0.45120671563483733, 'ood_validation': 0.5518867924528302, 'train': 0.9212483224120128}
Train Datashifts with Representations:{'informativeness': 0.16985550220841403, 'invariance': 1.1736813133351025, 'label_shift': 0.02660028896165906, '(latent)_covariate_shift': 1.1645609463101294, 'concept_shift': 0.080577044407797, 'residual': 0.09397155408916137, 'overall': 0.04433538688369719}
Test Datashifts with Representations:{'informativeness': 0.2542348690346613, 'invariance': 1.8706553092488023, 'label_shift': 0.02329921323212632, '(latent)_covariate_shift': 1.890871866798543, 'concept_shift': 0.05715790844597368, 'residual': 0.28530400719873195, 'overall': -0.037890206996124}
{'assistments_group_dro'}
{'validation': 0.894210064299273, 'id_test': 0.8951704268361307, 'ood_test': 0.4491080797481637, 'ood_validation': 0.5518867924528302, 'train': 0.8940158291153308}
Train Datashifts with Representations:{'informativeness': 0.15995815709897593, 'invariance': 1.0261882282039465, 'label_shift': 0.02660028896165906, '(latent)_covariate_shift': 1.0168128823897102, 'concept_shift': 0.0030443914271028123, 'residual': 0.12472036315619564, 'overall': 0.042328069802940294}
Test Datashifts with Representations:{'informativeness': 0.22785257547304774, 'invariance': 1.8748222699369799, 'label_shift': 0.02329921323212632, '(latent)_covariate_shift': 1.9033754614681924, 'concept_shift': -0.00912996418085792, 'residual': 0.28501309993628265, 'overall': -0.026669339991136576}
{'acspubcov_causal_group_dro'}
{'validation': 0.7770881882499536, 'id_test': 0.7762120044250792, 'ood_test': 0.36371728267208886, 'ood_validation': 0.3601500946344469, 'train': 0.7760952951251907}
Train Datashifts with Representations:{'informativeness': 0.02257457935508308, 'invariance': 0.0009955386106988574, 'label_shift': 0.016856964365221407, '(latent)_covariate_shift': -0.0009763673314429546, 'concept_shift': 0.015666115738971254, 'residual': 0.024070897183493513, 'overall': -0.0018868464863562619}
Test Datashifts with Representations:{'informativeness': 0.05010416802529054, 'invariance': -8.625282307065914e-05, 'label_shift': -0.003372463110508083, '(latent)_covariate_shift': 0.0005694352436983902, 'concept_shift': 0.11402648947253523, 'residual': 0.051995847192757946, 'overall': -0.06026331142560454}
{'acspubcov_arguablycausal_group_dro'}
{'validation': 0.800040336993616, 'id_test': 0.7990303165848612, 'ood_test': 0.5828345827062016, 'ood_validation': 0.5837294775298209, 'train': 0.7990876253572856}
Train Datashifts with Representations:{'informativeness': 0.09232870701358488, 'invariance': 0.036640713847399614, 'label_shift': 0.016856964365221407, '(latent)_covariate_shift': 0.00012142219362211753, 'concept_shift': 0.06327338980027927, 'residual': 0.06746181338009699, 'overall': -0.016600964910929786}
Test Datashifts with Representations:{'informativeness': 0.22120777449821832, 'invariance': 0.03790944688607927, 'label_shift': -0.003372463110508083, '(latent)_covariate_shift': 1.0367668448506073e-05, 'concept_shift': 0.0168110039639589, 'residual': 0.16059010186723233, 'overall': 0.03157639948493712}
{'acspubcov_group_dro'}
{'validation': 0.8084771586781447, 'id_test': 0.8077886186005088, 'ood_test': 0.6302512480483007, 'ood_validation': 0.630045336502487, 'train': 0.8082181112556908}
Train Datashifts with Representations:{'informativeness': 0.020511906720672837, 'invariance': 0.026293117120201426, 'label_shift': 0.016856964365221407, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.0019476383052003667, 'residual': 0.01689824110331461, 'overall': -0.0020782299127319673}
Test Datashifts with Representations:{'informativeness': 0.03482220839173183, 'invariance': 0.02120761710106815, 'label_shift': -0.003372463110508083, '(latent)_covariate_shift': 0.0, 'concept_shift': 0.0027411205776893695, 'residual': 0.02645271266821907, 'overall': -0.005291104671120043}
{'brfss_blood_pressure_causal_group_dro'}
{'validation': 0.5977967542789545, 'id_test': 0.5980334171225787, 'ood_test': 0.4157073938244039, 'ood_validation': 0.4164859002169197, 'train': 0.5987865681504175}
Train Datashifts with Representations:{'informativeness': 0.09965932863113837, 'invariance': 0.0692903686143497, 'label_shift': 0.01572578048184541, '(latent)_covariate_shift': 0.0663888171217777, 'concept_shift': 0.08139271642511489, 'residual': 0.09590285940196513, 'overall': -0.03052777448874751}
Test Datashifts with Representations:{'informativeness': 0.07104750201759404, 'invariance': 0.0391859029351735, 'label_shift': 0.02105310105154427, '(latent)_covariate_shift': 0.045352403637596087, 'concept_shift': 0.07502280913404194, 'residual': 0.080297887404611, 'overall': -0.03315198907705451}
{'brfss_blood_pressure_arguablycausal_group_dro'}
{'validation': 0.5977967542789545, 'id_test': 0.5980334171225787, 'ood_test': 0.4157073938244039, 'ood_validation': 0.4164859002169197, 'train': 0.5987865681504175}
Train Datashifts with Representations:{'informativeness': 0.09765699089084444, 'invariance': 0.07793818476605491, 'label_shift': 0.01572578048184541, '(latent)_covariate_shift': 0.0810877103026729, 'concept_shift': 0.13268393481321883, 'residual': 0.09644076972020636, 'overall': -0.055688093226739635}
Test Datashifts with Representations:{'informativeness': 0.06936658464312632, 'invariance': 0.04740045829685339, 'label_shift': 0.02105310105154427, '(latent)_covariate_shift': 0.055227487913560795, 'concept_shift': 0.12864109945946633, 'residual': 0.07159031959849071, 'overall': -0.05210421935097172}
{'brfss_blood_pressure_group_dro'}
{'validation': 0.5977967542789545, 'id_test': 0.5980334171225787, 'ood_test': 0.4157073938244039, 'ood_validation': 0.4164859002169197, 'train': 0.5987865681504175}
Train Datashifts with Representations:{'informativeness': 0.0986593699492623, 'invariance': 0.07989107455116866, 'label_shift': 0.01572578048184541, '(latent)_covariate_shift': 0.0806026583173446, 'concept_shift': 0.016078243855003775, 'residual': 0.09844055941794555, 'overall': 0.00039837072782553107}
Test Datashifts with Representations:{'informativeness': 0.07365563705977853, 'invariance': 0.05030459553382311, 'label_shift': 0.02105310105154427, '(latent)_covariate_shift': 0.05353167234564843, 'concept_shift': 0.01740055705282833, 'residual': 0.07587105035350622, 'overall': 0.0012243971115429297}
