Upload 5 files

NCD_synthetic.csv

@@ -0,0 +1,501 @@
+Id,AGE,DM_key,IFG_key,DM_ICD10,DM_drugs,Glucose,HbA1c,eGFR,HTN_key,cvd_key,HTN_ICD10,HTN_drugs,sbp1,dbp1,CKD_key,CKD_ICD10,CKD_drugs,DLP_key,DLP_ICD10,DLP_drugs,LDL-c,DM_label,HTN_label,CKD_label,DLP_label
+0,57,0,0,0,1,265,13,40,0,0,0,0,150,50,0,0,0,1,0,1,42,0,0,0,1
+1,35,1,0,1,1,210,7,52,0,0,0,0,100,80,0,0,0,0,0,0,70,1,0,0,0
+2,55,0,0,0,1,276,15,18,0,1,0,0,140,40,0,0,0,0,1,0,126,0,0,0,0
+3,74,0,1,0,1,14,16,86,1,1,1,0,120,70,0,0,1,0,0,0,162,0,1,0,0
+4,16,1,1,1,1,141,11,53,0,1,0,0,130,80,0,0,1,0,0,0,38,1,0,0,0
+5,17,0,0,0,1,216,8,18,0,0,0,0,100,70,0,0,0,0,0,0,13,0,0,0,0
+6,69,0,1,0,0,75,7,101,0,1,0,0,150,90,0,0,1,0,0,0,31,0,0,0,0
+7,40,0,0,0,1,79,10,83,0,0,0,0,150,60,0,0,1,0,0,0,171,0,0,0,0
+8,68,0,1,0,1,230,12,9,0,0,0,0,140,60,0,0,0,0,0,0,89,0,0,0,0
+9,80,0,1,0,0,104,14,46,0,1,0,1,140,70,0,0,0,0,0,0,94,0,0,0,0
+10,86,0,0,0,0,273,2,89,0,1,0,0,130,80,0,0,0,0,0,1,49,0,0,0,1
+11,11,0,0,0,0,6,2,20,0,1,0,0,80,70,0,1,1,0,0,1,175,0,0,0,1
+12,41,0,1,0,0,99,6,117,1,0,1,1,170,90,0,0,0,0,0,0,140,0,1,0,0
+13,21,0,0,0,1,77,7,23,0,1,1,0,130,90,0,0,0,0,0,0,118,0,1,0,0
+14,87,0,1,1,1,231,3,40,0,1,0,0,100,90,0,0,0,0,0,0,60,0,0,0,0
+15,41,0,1,0,0,56,0,49,0,1,0,0,150,80,0,0,0,0,0,1,44,0,0,0,0
+16,87,1,0,1,0,294,4,36,0,0,0,0,90,70,0,0,0,0,0,0,131,1,0,0,0
+17,68,0,0,0,0,281,14,48,0,0,0,0,130,60,1,0,1,0,0,0,157,0,0,1,0
+18,30,0,1,1,1,257,11,45,0,0,0,0,130,60,1,0,1,0,0,0,30,0,0,1,0
+19,23,0,0,0,0,68,4,100,0,0,1,0,120,70,0,0,0,0,0,0,20,0,1,0,0
+20,68,0,0,0,0,170,17,66,0,1,0,0,100,90,0,0,0,0,0,0,0,0,0,0,0
+21,34,0,1,1,0,149,17,35,0,1,0,0,140,60,0,0,0,0,0,0,123,0,0,0,0
+22,40,0,0,0,0,298,12,102,0,0,0,0,120,70,1,0,1,0,0,0,7,0,0,1,0
+23,48,0,0,1,0,20,10,8,0,1,0,0,150,40,0,0,1,0,0,0,54,0,0,0,0
+24,48,1,1,1,0,292,8,32,0,1,0,0,80,80,0,1,1,0,0,0,25,1,0,0,0
+25,83,0,1,0,0,206,5,34,0,1,0,0,130,90,0,0,0,0,0,0,183,0,0,0,0
+26,24,0,1,0,1,141,8,60,0,1,0,0,100,60,0,0,0,0,1,0,125,0,0,0,0
+27,12,1,0,1,0,26,1,19,0,1,0,0,150,80,0,0,0,0,1,1,50,1,0,0,0
+28,60,0,0,0,0,105,11,61,0,0,0,0,100,80,0,0,0,0,0,0,126,0,0,0,0
+29,28,0,1,0,0,318,13,88,0,0,0,0,80,60,0,0,0,0,0,0,8,0,0,0,0
+30,17,0,0,0,0,214,8,10,0,0,0,0,140,70,0,0,0,0,0,0,35,0,0,0,0
+31,85,0,1,0,0,219,10,21,0,1,0,0,120,70,0,0,0,0,0,0,133,0,0,0,0
+32,78,0,0,0,0,118,11,110,0,1,0,0,100,70,0,0,1,0,0,1,81,0,0,0,0
+33,50,0,1,0,0,71,13,12,0,1,0,0,80,90,0,0,0,0,0,0,64,0,0,0,0
+34,33,0,1,1,0,299,15,114,1,1,1,1,100,90,0,0,0,1,0,1,61,0,1,0,1
+35,59,1,0,1,0,17,15,121,0,0,0,0,110,80,0,0,0,1,0,1,15,1,0,0,1
+36,16,0,1,0,0,205,15,98,0,1,0,1,120,90,0,1,1,0,0,0,26,0,0,0,0
+37,4,0,1,0,0,57,7,64,0,1,0,1,90,90,0,0,0,1,0,1,63,0,0,0,1
+38,14,0,0,0,1,51,5,126,0,0,0,0,130,90,0,0,0,0,0,0,111,0,0,0,0
+39,18,0,0,0,0,140,4,22,0,0,0,0,100,90,0,0,0,0,0,0,101,0,0,0,0
+40,39,0,1,0,1,125,0,32,0,0,0,0,110,80,0,0,1,0,0,0,116,0,0,0,0
+41,64,0,0,0,0,92,15,114,0,1,0,0,140,90,0,0,0,0,0,0,24,0,0,0,0
+42,86,0,1,0,0,189,14,48,0,0,0,0,120,60,0,0,0,1,0,1,120,0,0,0,1
+43,5,0,0,0,0,241,2,42,0,1,0,0,130,70,0,0,0,0,0,0,108,0,0,0,0
+44,68,0,1,1,0,281,6,105,1,0,1,0,120,90,0,1,0,0,0,0,150,0,1,0,0
+45,50,0,1,0,0,295,4,34,0,1,0,0,120,80,0,0,0,0,0,1,130,0,0,0,0
+46,71,0,0,1,0,85,8,78,0,0,0,0,140,80,1,1,1,0,0,0,28,0,0,1,0
+47,67,0,0,0,0,250,15,87,0,1,1,0,130,70,1,0,1,0,0,0,85,0,1,1,0
+48,3,0,1,0,0,22,7,38,0,0,0,0,110,80,0,0,0,0,0,1,77,0,0,0,0
+49,71,0,1,0,1,214,3,26,0,0,0,0,140,60,1,1,1,0,0,0,32,0,0,1,0
+50,73,0,0,0,0,135,0,108,0,0,0,0,100,80,0,0,0,0,0,0,79,0,0,0,0
+51,71,0,0,0,0,168,8,84,0,0,0,0,110,90,0,0,0,0,0,1,124,0,0,0,1
+52,28,0,1,0,0,150,17,116,0,0,0,0,80,70,0,0,0,0,0,0,140,0,0,0,0
+53,30,0,1,0,0,261,10,30,0,0,1,1,130,90,0,0,0,0,0,1,132,0,1,0,0
+54,50,0,0,0,0,253,10,57,1,1,0,0,130,90,0,0,0,0,0,0,11,0,1,0,0
+55,29,0,1,0,0,84,7,125,0,1,0,0,80,90,0,0,1,0,0,0,23,0,0,0,0
+56,15,0,1,0,1,52,2,113,0,0,0,0,150,60,0,0,0,0,0,0,58,0,0,0,0
+57,5,0,0,0,0,247,16,129,0,1,0,1,150,60,0,0,0,0,0,0,45,0,0,0,0
+58,56,0,0,0,0,206,16,107,0,0,0,0,100,70,0,0,0,0,0,0,77,0,0,0,0
+59,22,0,0,0,0,220,0,20,0,1,0,0,130,90,0,0,0,0,0,0,161,0,0,0,0
+60,43,0,1,0,1,164,16,85,0,1,0,1,100,90,0,0,0,1,0,1,172,0,1,0,1
+61,2,0,1,0,0,284,2,105,0,0,0,0,130,60,1,1,1,0,0,0,61,0,0,1,0
+62,38,0,0,0,0,289,8,35,0,0,0,0,110,50,0,0,0,0,0,0,166,0,0,0,0
+63,11,0,0,0,1,0,16,92,0,1,0,0,120,70,0,0,0,0,0,0,154,0,0,0,0
+64,31,0,0,0,0,268,12,0,1,0,1,0,120,70,0,1,0,0,0,0,180,0,1,0,0
+65,6,0,1,0,0,201,11,65,0,0,0,0,120,80,0,0,1,1,0,1,88,0,0,0,1
+66,66,0,1,0,0,126,8,8,0,1,0,0,120,80,1,1,1,0,0,0,85,0,0,1,0
+67,84,1,0,1,0,2,10,70,0,0,0,1,100,90,0,0,0,1,0,1,181,1,0,0,1
+68,15,0,1,0,1,162,2,115,0,1,0,0,120,60,0,0,0,0,0,0,75,0,0,0,0
+69,10,0,1,0,1,182,6,52,0,0,0,0,140,90,0,0,0,0,0,0,93,0,0,0,0
+70,55,0,0,0,0,111,13,110,0,0,0,0,130,90,0,0,0,0,0,0,109,0,0,0,0
+71,69,0,1,0,1,288,11,24,0,0,0,0,140,80,0,0,1,0,0,1,116,0,0,0,1
+72,33,0,1,0,0,207,13,82,0,1,0,0,100,80,0,0,1,0,0,0,12,0,0,0,0
+73,17,1,1,1,1,60,12,100,0,0,0,0,130,70,0,0,1,0,0,0,33,1,1,0,0
+74,82,0,0,0,0,113,2,111,0,1,0,0,130,70,0,0,0,0,0,0,90,0,0,0,0
+75,50,0,1,0,0,61,8,54,0,0,0,0,110,70,0,1,0,0,0,0,64,0,0,0,0
+76,44,0,0,0,0,134,14,41,0,1,0,0,110,80,0,0,1,0,0,1,142,0,0,0,0
+77,56,0,0,0,0,61,0,21,0,1,0,1,100,40,0,0,0,0,0,0,169,0,0,0,0
+78,15,1,1,1,1,277,10,76,0,0,0,1,130,90,0,0,0,0,0,0,100,1,1,0,0
+79,13,0,1,0,0,114,7,33,0,0,0,0,130,70,0,0,1,0,0,0,30,0,0,0,0
+80,10,0,1,0,0,205,9,5,0,0,0,0,140,60,0,0,0,0,0,1,189,0,0,0,0
+81,10,0,1,0,0,274,17,26,0,1,0,0,100,60,0,0,0,1,0,1,62,0,0,0,1
+82,22,0,0,0,0,153,4,12,0,1,0,0,110,60,0,0,0,0,0,1,72,0,0,0,0
+83,27,0,1,0,0,208,11,19,0,0,0,0,80,80,0,0,0,0,0,0,136,0,0,0,0
+84,35,0,0,0,0,23,9,86,0,0,0,0,120,80,0,0,1,0,0,0,56,0,0,0,0
+85,33,0,0,1,0,212,14,57,0,1,1,0,130,90,0,0,0,0,0,0,120,0,1,0,0
+86,20,0,0,0,0,142,1,111,0,0,0,0,120,80,0,1,0,1,0,1,124,0,0,0,1
+87,7,0,0,0,0,275,6,72,0,1,0,1,140,90,0,0,0,0,0,0,171,0,0,0,0
+88,15,0,0,0,0,26,11,108,0,1,0,0,110,80,0,1,1,0,0,0,153,0,0,0,0
+89,86,0,1,0,0,192,6,6,0,0,0,0,130,60,0,0,0,0,0,0,116,0,0,0,0
+90,40,0,0,0,0,200,15,94,0,1,0,1,110,80,0,0,1,0,0,0,165,0,0,0,0
+91,25,0,0,0,0,139,11,50,1,1,0,0,130,90,0,0,0,0,0,0,179,0,1,0,0
+92,62,0,0,0,0,312,7,43,0,0,0,1,130,90,0,0,1,0,0,0,5,0,1,0,0
+93,12,0,1,0,0,10,13,14,0,0,0,0,140,50,0,0,1,0,0,0,6,0,0,0,0
+94,25,0,1,0,1,187,14,70,0,0,0,0,100,70,0,0,1,0,0,0,144,0,1,0,0
+95,53,0,1,0,0,85,6,84,0,0,0,0,130,70,0,0,0,0,0,0,169,0,0,0,0
+96,70,0,0,0,0,121,12,13,0,1,0,0,140,70,0,0,1,0,0,0,23,0,0,0,0
+97,54,0,1,0,0,197,1,24,0,0,0,0,110,40,0,0,0,1,0,1,107,0,0,0,1
+98,58,0,0,0,0,197,16,32,0,0,0,0,150,90,0,0,0,0,0,0,63,0,0,0,0
+99,31,0,0,0,0,167,14,111,0,0,0,0,110,50,0,0,0,0,0,0,55,0,0,0,0
+100,65,0,1,0,0,204,12,70,0,1,0,0,130,90,0,0,1,0,0,1,75,,,,
+101,46,0,0,0,1,41,17,121,0,0,0,0,150,60,0,0,0,0,0,0,120,,,,
+102,75,0,0,0,0,191,8,121,1,0,0,0,120,90,0,0,0,1,0,1,119,,,,
+103,27,0,0,0,0,52,13,31,0,1,0,0,130,60,0,0,1,0,0,1,128,,,,
+104,12,1,1,1,1,158,3,65,0,1,0,0,110,60,0,1,0,0,0,0,29,,,,
+105,33,0,0,0,1,136,10,81,1,1,1,0,120,90,0,0,0,0,0,0,122,,,,
+106,24,0,1,0,1,209,14,106,0,0,0,0,130,80,0,0,0,0,0,0,109,,,,
+107,13,0,0,0,0,70,4,63,0,0,0,0,110,90,0,0,0,0,0,1,187,,,,
+108,64,0,0,0,0,81,1,109,0,0,0,0,80,50,0,1,0,0,0,0,37,,,,
+109,5,0,0,0,0,257,7,110,0,0,0,0,120,80,0,1,0,0,0,1,188,,,,
+110,60,0,0,0,0,263,9,29,0,1,0,0,140,80,0,0,0,0,0,0,137,,,,
+111,36,0,1,0,0,82,3,121,0,1,0,0,120,50,0,0,0,0,0,0,41,,,,
+112,49,0,0,0,1,65,8,124,0,1,0,0,100,40,1,1,1,0,0,0,156,,,,
+113,77,0,1,0,0,135,1,24,0,1,0,0,110,80,0,0,1,0,0,0,15,,,,
+114,58,0,0,0,0,217,2,33,0,0,0,0,110,90,0,0,0,0,0,1,74,,,,
+115,36,0,0,0,0,86,1,120,0,1,0,0,120,60,0,0,0,0,1,0,85,,,,
+116,4,0,1,0,0,95,6,96,0,0,0,0,110,60,0,0,0,0,0,0,173,,,,
+117,60,0,0,0,0,305,15,122,0,0,0,0,90,80,0,0,0,0,0,1,71,,,,
+118,36,0,0,0,0,307,6,79,0,1,0,0,100,90,0,0,0,0,0,0,37,,,,
+119,74,0,0,1,1,66,15,92,0,1,0,0,100,80,0,0,0,0,0,0,11,,,,
+120,23,1,0,1,1,269,17,64,0,1,0,0,120,70,0,0,1,0,0,0,75,,,,
+121,44,0,0,0,0,45,17,113,0,0,0,1,110,80,0,0,0,0,0,1,14,,,,
+122,81,0,0,1,0,97,11,47,1,0,1,1,120,70,0,0,0,0,0,0,167,,,,
+123,37,0,1,0,1,53,2,103,0,1,0,0,150,80,0,0,1,0,0,0,55,,,,
+124,5,0,0,0,0,133,13,73,0,1,0,0,130,80,0,0,0,0,0,0,98,,,,
+125,48,0,1,0,0,231,12,123,0,0,0,0,100,70,0,0,0,0,0,0,139,,,,
+126,31,0,0,0,0,267,9,83,0,1,0,0,120,90,0,0,1,0,0,0,135,,,,
+127,25,0,1,0,1,46,17,5,0,1,0,0,150,60,1,1,1,0,0,0,163,,,,
+128,17,1,1,1,1,223,5,32,1,0,0,1,120,90,0,0,1,0,0,0,36,,,,
+129,49,0,1,0,0,226,9,51,0,0,0,0,150,90,0,0,1,0,0,0,162,,,,
+130,70,1,1,1,1,315,9,2,0,0,0,0,140,70,0,0,0,0,0,1,21,,,,
+131,1,0,1,0,0,32,16,59,0,1,0,0,120,60,0,0,1,0,0,0,111,,,,
+132,59,0,0,0,1,251,5,116,0,0,0,0,100,70,0,0,1,0,0,0,96,,,,
+133,82,1,1,1,0,42,2,85,0,0,0,0,120,70,0,0,0,0,0,1,5,,,,
+134,55,0,0,0,0,211,14,119,0,0,0,1,120,90,0,0,0,0,0,0,80,,,,
+135,46,0,1,1,0,171,5,5,0,0,0,1,140,80,0,0,0,1,0,1,47,,,,
+136,66,0,0,0,0,291,1,96,0,1,0,0,100,40,0,0,0,0,0,0,104,,,,
+137,65,0,1,0,0,168,14,65,0,1,0,0,150,70,0,0,0,0,0,0,192,,,,
+138,38,0,1,0,0,202,10,71,0,1,0,0,120,70,0,0,0,0,1,0,84,,,,
+139,58,0,0,0,1,296,4,57,0,0,0,0,110,90,0,0,0,0,0,0,0,,,,
+140,37,0,0,0,0,78,1,52,0,1,0,0,80,90,0,0,0,0,0,0,108,,,,
+141,20,0,1,0,0,18,8,75,0,1,0,0,100,80,0,0,0,0,0,0,78,,,,
+142,42,0,1,0,1,31,4,86,0,0,0,0,80,80,0,0,0,0,0,0,80,,,,
+143,77,0,0,0,0,112,13,117,0,0,0,0,80,80,0,0,0,0,0,0,45,,,,
+144,16,0,0,0,0,139,8,50,0,1,0,0,150,90,0,0,0,0,0,0,51,,,,
+145,34,0,0,0,0,1,17,54,0,0,0,0,140,90,0,0,0,0,0,1,19,,,,
+146,22,1,0,1,1,112,3,94,1,1,0,0,130,70,0,0,0,0,0,0,51,,,,
+147,6,0,1,0,0,270,17,68,1,1,1,0,120,90,0,0,0,0,0,0,172,,,,
+148,34,0,1,0,0,292,12,15,0,1,0,0,100,60,0,0,1,0,0,0,49,,,,
+149,49,0,1,0,0,154,13,123,1,0,0,0,130,90,0,0,0,0,0,1,96,,,,
+150,28,0,1,0,1,169,0,103,1,1,1,1,120,70,0,0,0,0,1,0,9,,,,
+151,47,0,1,0,0,72,6,2,0,0,0,0,150,80,0,0,0,0,0,0,73,,,,
+152,21,0,1,0,0,35,5,15,0,0,0,0,140,70,0,0,0,1,0,1,190,,,,
+153,52,1,1,1,1,100,8,95,0,0,0,0,150,80,0,0,0,0,0,0,149,,,,
+154,1,0,0,0,0,270,0,3,0,1,0,0,100,50,0,0,0,0,0,0,16,,,,
+155,32,0,0,0,0,196,6,46,0,1,0,0,120,60,0,0,1,0,0,1,134,,,,
+156,33,0,0,0,0,43,13,38,0,0,0,1,130,80,0,0,0,0,0,0,20,,,,
+157,6,0,0,0,0,200,16,49,0,1,0,0,120,80,0,0,0,0,1,0,114,,,,
+158,67,0,0,0,0,310,17,58,0,0,0,0,130,60,0,1,1,0,1,0,167,,,,
+159,79,0,0,0,0,166,14,125,0,1,0,0,100,70,0,0,0,0,0,0,113,,,,
+160,20,0,1,0,0,173,10,101,0,0,0,0,130,60,0,0,0,0,0,0,2,,,,
+161,21,0,1,0,1,249,4,123,0,1,0,0,130,80,0,0,0,0,0,0,62,,,,
+162,65,0,0,0,0,130,5,41,0,0,0,0,140,50,0,1,0,0,0,0,175,,,,
+163,30,1,1,1,0,240,3,63,1,0,0,1,120,90,0,0,1,0,1,0,42,,,,
+164,45,0,1,0,0,199,2,42,0,1,0,0,150,90,0,1,0,0,0,0,34,,,,
+165,14,0,0,0,0,313,7,39,0,1,0,0,100,70,0,0,1,0,0,0,57,,,,
+166,80,0,0,0,0,178,10,60,0,0,0,0,130,70,0,0,0,0,0,0,94,,,,
+167,28,0,0,0,1,42,12,51,0,1,0,0,120,80,0,0,0,0,0,0,161,,,,
+168,54,0,0,1,1,156,8,17,0,1,0,0,150,80,0,0,0,0,0,0,40,,,,
+169,56,0,0,0,0,122,13,119,0,0,0,0,150,40,1,0,1,0,0,0,47,,,,
+170,64,0,1,0,0,8,11,21,0,1,0,1,120,80,0,0,0,0,0,0,156,,,,
+171,47,0,1,0,0,299,12,92,0,1,0,0,120,90,0,0,0,0,0,0,170,,,,
+172,3,1,1,1,1,308,0,40,0,1,0,0,120,40,0,0,0,0,0,1,143,,,,
+173,43,0,0,0,0,93,7,99,0,1,0,0,130,70,0,0,0,0,0,0,95,,,,
+174,37,0,1,0,0,19,9,3,0,0,0,0,80,40,0,0,0,1,0,1,117,,,,
+175,43,0,0,0,1,115,5,16,0,0,0,0,140,90,0,0,0,0,0,1,159,,,,
+176,34,0,1,0,0,318,15,17,0,1,1,0,120,90,0,0,0,0,0,1,93,,,,
+177,59,0,1,0,0,225,9,62,0,1,0,0,150,80,0,0,1,0,0,0,92,,,,
+178,31,0,0,0,0,317,9,25,0,0,0,0,120,80,0,0,0,0,0,0,62,,,,
+179,45,0,1,0,0,66,6,55,0,0,0,0,100,60,0,1,1,0,0,1,17,,,,
+180,7,0,0,0,0,58,11,4,1,1,1,1,100,90,0,0,0,0,0,0,112,,,,
+181,10,0,1,1,0,236,7,52,0,1,0,0,120,50,0,0,0,0,0,0,97,,,,
+182,23,0,1,0,0,176,9,87,0,1,0,0,120,70,0,0,0,0,0,0,180,,,,
+183,42,0,1,0,0,74,5,102,1,1,0,1,120,70,0,0,0,0,0,0,148,,,,
+184,87,0,0,0,0,45,3,112,0,0,0,0,120,40,0,0,0,0,0,0,87,,,,
+185,67,0,0,0,0,48,2,77,0,0,0,0,140,60,0,0,0,0,0,0,164,,,,
+186,56,0,1,0,0,44,2,92,1,1,0,1,120,70,1,0,1,0,0,0,42,,,,
+187,12,0,1,0,0,227,13,88,0,1,0,0,130,50,0,0,1,0,0,1,68,,,,
+188,74,0,0,0,0,101,4,74,0,0,0,0,80,80,0,0,0,0,0,0,128,,,,
+189,50,0,1,0,0,69,4,56,0,1,0,0,130,80,0,0,1,0,0,0,178,,,,
+190,7,0,0,0,0,244,13,74,0,1,0,0,110,80,0,0,0,0,0,0,46,,,,
+191,66,1,1,1,1,98,15,45,0,1,0,0,130,40,0,0,1,0,0,0,55,,,,
+192,17,0,1,0,0,264,7,85,0,0,0,0,110,90,1,0,1,0,1,0,187,,,,
+193,13,0,1,0,0,242,11,83,0,1,0,0,120,40,0,0,1,0,0,0,63,,,,
+194,40,0,0,0,1,266,15,25,0,0,0,0,130,40,0,0,0,0,0,0,43,,,,
+195,3,0,1,0,1,133,15,15,0,1,0,0,120,80,0,0,0,1,0,1,182,,,,
+196,52,0,0,0,0,140,0,6,0,1,0,0,100,60,0,0,0,0,0,0,12,,,,
+197,4,0,1,0,0,304,15,71,0,0,0,0,150,80,0,0,0,0,0,0,58,,,,
+198,48,0,0,0,0,314,0,78,0,0,0,0,120,50,0,0,0,0,0,0,10,,,,
+199,23,1,1,1,1,284,12,88,0,1,0,0,150,70,0,1,1,0,0,0,39,,,,
+200,26,0,1,0,0,145,9,7,1,0,1,0,100,70,0,0,1,0,0,1,176,,,,
+201,41,0,0,0,1,137,5,28,1,0,0,1,130,90,0,0,0,0,0,1,82,,,,
+202,11,0,1,0,0,40,2,33,0,1,0,0,110,80,0,0,0,0,0,0,123,,,,
+203,72,0,0,0,0,254,3,112,0,0,0,0,130,90,0,0,0,0,0,0,184,,,,
+204,70,0,0,0,0,12,10,54,1,0,0,0,120,90,0,0,0,0,0,0,178,,,,
+205,59,0,0,1,0,278,4,101,0,1,0,0,140,80,1,0,1,1,0,1,68,,,,
+206,36,0,0,0,0,179,15,105,0,1,0,0,120,90,0,0,0,0,0,0,3,,,,
+207,12,0,1,0,0,238,8,63,1,1,1,1,130,90,0,0,0,0,0,0,71,,,,
+208,83,0,0,0,0,234,1,53,1,1,0,1,130,90,0,0,0,0,0,0,99,,,,
+209,63,0,0,0,0,227,4,106,0,1,0,0,140,70,0,0,1,0,1,0,52,,,,
+210,6,0,0,0,0,291,1,73,0,0,0,1,110,60,0,0,0,0,0,1,131,,,,
+211,8,1,0,1,1,204,1,79,0,0,0,0,90,80,0,0,0,0,0,0,69,,,,
+212,87,0,0,0,1,11,12,115,0,1,0,0,120,60,0,0,0,0,1,1,152,,,,
+213,57,0,0,0,0,130,11,22,0,1,0,0,130,80,0,0,1,0,0,0,47,,,,
+214,9,0,0,0,0,233,9,128,0,0,0,0,150,80,0,0,0,0,0,0,181,,,,
+215,44,0,1,0,0,193,14,69,0,1,0,0,100,90,0,0,0,0,0,0,190,,,,
+216,63,0,0,0,0,192,12,100,0,1,0,0,100,70,0,0,0,0,0,0,46,,,,
+217,24,0,1,0,0,110,0,79,0,0,0,0,110,90,0,0,0,1,0,1,49,,,,
+218,39,0,1,0,0,263,11,45,0,1,0,0,140,40,0,0,0,0,0,0,143,,,,
+219,74,0,0,0,0,315,16,119,0,1,0,1,100,60,0,0,1,0,1,0,102,,,,
+220,26,0,1,0,0,245,12,96,0,0,0,0,90,50,0,0,1,0,0,0,194,,,,
+221,15,0,1,0,0,283,3,72,0,1,0,0,80,70,0,0,1,0,0,0,78,,,,
+222,65,0,1,0,0,172,11,70,0,1,0,0,130,60,0,0,1,0,0,0,82,,,,
+223,11,0,0,0,1,184,17,67,0,1,0,0,150,80,0,1,1,0,0,0,132,,,,
+224,25,0,0,0,0,97,1,47,0,1,0,0,150,70,1,1,1,0,1,0,160,,,,
+225,78,0,1,0,0,317,9,102,0,0,0,0,140,90,0,0,0,0,0,0,177,,,,
+226,35,0,1,0,0,267,3,73,0,0,0,0,120,60,0,0,0,0,0,0,135,,,,
+227,61,0,1,0,0,93,12,54,0,0,0,0,130,70,0,0,0,0,0,0,8,,,,
+228,8,0,1,0,1,33,9,0,1,1,1,1,120,90,0,0,1,0,0,0,173,,,,
+229,54,0,1,0,0,136,9,127,0,0,0,0,150,80,0,0,0,0,0,1,162,,,,
+230,47,0,0,0,1,304,6,126,0,0,0,0,130,80,0,0,0,0,0,0,102,,,,
+231,29,0,1,0,0,175,16,76,0,0,0,0,100,80,0,0,0,0,0,0,7,,,,
+232,15,0,0,0,0,132,17,0,0,1,0,0,150,70,1,1,1,0,1,0,6,,,,
+233,67,0,1,0,1,132,5,29,0,0,0,0,150,80,0,0,0,0,0,0,122,,,,
+234,75,0,0,0,0,148,15,81,1,1,0,1,130,90,0,0,1,0,0,0,87,,,,
+235,52,0,0,0,0,119,11,109,0,1,0,0,130,80,0,0,0,0,0,0,91,,,,
+236,65,0,0,0,0,180,2,14,0,0,0,0,100,90,0,0,0,0,0,0,137,,,,
+237,32,0,0,0,0,159,0,106,0,1,0,0,90,50,0,0,0,0,1,1,114,,,,
+238,26,0,0,0,0,43,2,22,0,1,0,0,150,40,0,0,1,0,0,0,117,,,,
+239,79,0,1,1,0,170,5,11,0,0,0,0,150,80,0,0,1,0,0,1,193,,,,
+240,39,0,1,0,0,89,17,116,0,0,0,0,120,40,0,0,0,0,1,1,48,,,,
+241,75,0,0,0,0,173,0,11,0,0,0,0,100,80,0,0,0,0,0,0,189,,,,
+242,1,0,0,0,0,111,3,42,0,1,0,0,100,70,0,0,0,0,0,1,141,,,,
+243,9,1,0,1,1,301,3,8,0,1,0,0,120,80,0,0,0,0,0,1,57,,,,
+244,19,1,1,1,1,282,14,38,0,1,0,0,110,90,0,0,0,0,0,0,188,,,,
+245,46,0,1,0,0,108,7,78,0,1,0,0,150,40,0,0,0,0,0,0,147,,,,
+246,69,0,1,0,0,31,17,61,0,0,0,0,140,50,0,0,0,0,0,0,112,,,,
+247,43,1,0,1,1,0,17,127,0,0,0,0,150,80,0,0,0,1,0,1,180,,,,
+248,24,0,1,0,1,152,12,98,0,1,0,0,120,50,0,0,1,0,0,0,163,,,,
+249,51,0,0,0,0,39,17,37,0,0,0,0,120,60,0,0,0,0,0,1,155,,,,
+250,38,0,0,0,0,94,1,104,0,1,0,1,150,90,0,0,0,0,0,0,176,,,,
+251,31,0,0,0,0,300,15,13,0,0,0,0,110,70,0,0,1,1,0,1,69,,,,
+252,20,0,1,0,0,24,14,90,0,0,0,1,150,70,0,0,0,0,0,0,1,,,,
+253,59,0,1,0,0,116,1,6,0,0,0,0,110,80,0,0,0,0,0,0,159,,,,
+254,32,0,1,0,0,183,11,117,0,1,0,0,140,40,0,0,0,0,0,0,60,,,,
+255,81,0,1,0,0,161,11,94,0,1,0,0,120,60,0,0,0,0,0,0,22,,,,
+256,7,0,0,0,1,44,13,25,0,1,0,1,80,90,0,0,0,0,0,1,36,,,,
+257,36,0,1,0,0,143,11,65,0,0,0,0,140,90,1,0,1,0,0,0,136,,,,
+258,72,0,1,0,1,280,9,25,0,0,0,0,120,90,0,0,1,0,0,0,170,,,,
+259,60,0,1,0,0,293,17,89,0,0,0,0,150,90,0,0,1,0,0,0,133,,,,
+260,77,0,1,0,0,232,1,60,0,0,0,0,110,80,0,0,1,0,0,0,193,,,,
+261,26,0,1,0,1,37,2,19,0,0,0,0,80,70,0,0,0,0,0,0,26,,,,
+262,3,0,0,0,0,228,6,82,0,0,0,0,120,60,0,0,0,0,0,0,124,,,,
+263,80,0,0,0,0,90,3,127,0,0,0,0,150,80,0,0,1,1,0,1,129,,,,
+264,5,0,1,0,0,75,10,105,1,1,1,0,130,70,0,0,0,0,0,0,113,,,,
+265,16,0,0,0,0,213,16,30,0,0,0,1,140,40,0,0,0,0,0,1,138,,,,
+266,46,0,0,0,0,102,5,43,0,1,0,0,80,90,0,0,0,0,1,0,3,,,,
+267,77,1,0,1,1,218,0,21,0,1,0,0,100,50,0,0,0,0,0,0,185,,,,
+268,57,0,0,0,0,17,3,51,0,0,0,0,100,60,0,0,0,0,0,0,194,,,,
+269,18,0,1,0,1,71,4,59,0,0,0,0,140,60,0,0,1,1,0,1,83,,,,
+270,4,1,1,1,0,55,0,126,0,0,0,1,120,80,0,0,0,0,0,1,106,,,,
+271,69,1,0,1,1,106,8,18,0,1,0,1,130,70,0,0,0,0,0,0,154,,,,
+272,42,0,0,0,0,245,12,89,0,0,0,0,140,80,1,1,1,0,0,0,13,,,,
+273,9,0,0,0,0,232,1,66,0,0,0,0,150,70,0,0,1,0,0,0,104,,,,
+274,73,0,0,0,0,91,5,87,0,1,0,0,140,70,0,0,0,0,0,0,65,,,,
+275,65,0,1,0,0,18,7,98,0,1,0,0,140,90,0,1,0,0,0,1,1,,,,
+276,85,0,1,0,0,190,1,11,0,1,0,0,150,90,0,0,0,0,0,1,27,,,,
+277,6,0,0,0,0,246,6,35,0,0,0,0,150,90,0,0,0,0,0,0,18,,,,
+278,77,1,0,1,0,184,6,75,0,1,1,1,120,90,0,0,0,0,0,0,152,,,,
+279,67,0,0,1,1,155,3,117,0,0,0,0,150,70,0,0,0,0,0,1,70,,,,
+280,76,1,0,1,0,282,7,37,0,0,0,0,140,80,0,0,0,0,0,0,4,,,,
+281,67,0,0,0,0,169,10,95,0,0,0,0,100,70,0,0,0,0,0,0,110,,,,
+282,40,0,1,0,0,37,12,57,0,0,0,0,140,90,0,0,0,0,0,0,191,,,,
+283,38,0,1,0,0,96,5,84,1,1,0,0,120,90,0,0,0,0,0,0,41,,,,
+284,19,0,1,1,0,30,1,34,0,0,0,0,140,90,0,0,0,0,0,1,72,,,,
+285,78,1,0,1,1,127,17,50,0,1,0,0,140,90,0,0,0,0,0,0,51,,,,
+286,26,0,1,0,0,256,2,73,0,0,0,1,100,70,0,0,0,0,0,0,167,,,,
+287,21,0,0,1,0,181,7,19,0,0,0,0,80,70,0,0,0,0,0,0,52,,,,
+288,66,0,1,0,0,244,14,84,0,0,0,0,140,80,0,0,0,0,0,1,16,,,,
+289,48,0,0,0,0,8,6,17,0,1,0,0,120,40,0,0,0,0,0,0,113,,,,
+290,32,0,0,0,0,57,8,55,0,1,0,0,150,70,0,0,0,0,0,0,27,,,,
+291,82,0,1,0,0,251,5,31,0,0,0,0,80,60,0,0,0,0,0,1,147,,,,
+292,62,0,1,0,0,222,13,9,0,1,0,0,140,40,0,0,1,0,0,0,100,,,,
+293,61,1,0,1,1,165,0,9,0,0,0,0,130,90,0,0,0,1,0,1,157,,,,
+294,80,0,0,0,0,67,17,66,1,0,0,0,130,90,0,0,0,0,0,0,103,,,,
+295,33,0,0,0,0,14,0,64,0,0,0,1,110,60,0,0,0,0,0,0,174,,,,
+296,29,0,1,0,0,229,9,95,0,1,0,0,110,80,0,0,0,0,0,1,27,,,,
+297,14,0,0,0,1,59,5,118,0,1,0,0,100,40,0,0,0,0,0,0,14,,,,
+298,53,0,1,0,0,233,3,39,0,1,0,0,100,70,0,0,0,0,0,0,4,,,,
+299,22,0,1,0,0,33,13,16,0,1,0,0,100,40,0,0,0,0,0,0,86,,,,
+300,8,0,1,0,0,210,0,24,1,1,1,1,120,90,0,0,0,0,0,0,171,,,,
+301,18,0,0,0,1,222,0,86,0,0,0,0,140,80,0,0,0,1,0,0,58,,,,
+302,45,0,0,0,0,9,15,79,0,1,0,0,140,60,0,0,0,1,0,1,156,,,,
+303,36,0,1,0,0,34,14,58,0,1,0,0,140,80,0,0,0,0,0,0,29,,,,
+304,62,0,1,0,1,80,13,71,1,0,0,0,130,90,0,0,0,0,0,1,19,,,,
+305,85,0,1,0,0,215,14,128,1,0,0,1,130,90,0,0,0,0,0,0,185,,,,
+306,83,0,0,0,0,117,0,33,0,0,0,0,150,70,0,0,0,0,0,0,35,,,,
+307,45,0,1,0,0,137,8,23,0,1,0,0,120,80,0,0,0,0,0,0,76,,,,
+308,63,0,1,0,0,221,3,37,0,0,0,0,140,60,0,0,1,0,0,1,70,,,,
+309,35,0,1,0,0,220,13,103,0,0,0,0,150,70,0,0,0,0,0,0,103,,,,
+310,75,1,1,1,0,286,14,128,1,0,1,0,100,70,0,0,0,0,0,0,95,,,,
+311,68,0,1,0,0,272,1,0,0,0,0,0,150,90,0,0,0,0,0,0,133,,,,
+312,43,0,0,0,0,186,3,125,0,1,1,0,130,90,0,0,0,0,0,0,158,,,,
+313,4,0,1,0,0,129,7,108,0,0,0,0,150,90,0,0,1,0,0,0,174,,,,
+314,49,1,0,1,1,302,8,27,0,0,0,0,120,70,0,0,0,1,0,1,105,,,,
+315,14,1,1,1,1,23,14,44,1,1,1,0,120,70,0,0,1,0,0,0,8,,,,
+316,16,0,1,0,1,285,3,6,0,0,0,0,130,60,0,0,0,0,0,0,185,,,,
+317,46,0,0,1,0,87,15,124,0,1,0,0,100,80,0,1,0,0,0,1,184,,,,
+318,39,0,0,0,0,56,6,120,0,1,0,0,110,70,0,0,1,0,0,0,145,,,,
+319,72,0,0,0,0,241,14,110,0,0,0,0,80,70,0,0,1,0,0,0,38,,,,
+320,64,0,0,0,1,146,6,75,0,0,0,1,120,70,0,0,0,0,0,0,53,,,,
+321,60,0,0,0,0,259,7,10,0,0,0,0,150,90,0,0,0,0,0,0,95,,,,
+322,0,0,1,0,0,58,8,104,0,1,0,0,130,80,1,1,1,0,0,0,184,,,,
+323,76,0,1,0,1,13,2,3,1,1,0,1,100,90,0,0,0,0,0,0,160,,,,
+324,23,0,0,0,0,90,10,46,0,0,0,0,150,60,0,0,1,0,0,0,139,,,,
+325,75,0,1,0,0,188,8,112,0,1,0,0,110,60,0,0,0,0,0,0,56,,,,
+326,37,0,0,0,0,114,6,104,0,1,0,0,140,60,0,0,1,0,0,0,97,,,,
+327,83,0,0,0,0,235,6,128,1,1,1,0,120,90,0,0,0,0,0,0,0,,,,
+328,30,0,0,0,0,249,3,38,0,1,0,0,110,80,0,0,0,0,0,0,59,,,,
+329,24,0,0,0,0,161,6,80,1,0,0,1,130,70,0,0,0,1,0,1,31,,,,
+330,80,0,1,0,0,223,10,58,0,1,0,0,110,60,0,0,0,0,0,0,90,,,,
+331,2,0,1,0,0,50,16,13,0,0,0,0,100,50,0,0,0,0,0,0,115,,,,
+332,12,0,0,1,0,143,13,107,0,0,0,0,90,60,0,0,1,0,0,1,64,,,,
+333,63,0,1,0,1,216,2,60,0,0,0,0,100,90,0,0,0,0,0,0,18,,,,
+334,54,0,0,0,0,309,17,93,0,1,0,0,110,80,0,0,0,0,1,0,157,,,,
+335,28,0,0,0,0,254,16,91,0,1,0,0,80,70,0,0,0,0,0,0,66,,,,
+336,66,0,0,0,0,269,14,97,0,1,0,0,80,40,0,0,0,0,1,0,174,,,,
+337,21,0,1,0,0,81,17,48,0,1,0,0,140,60,0,0,1,0,0,0,50,,,,
+338,78,0,1,0,0,295,1,109,0,1,0,0,100,60,1,1,1,0,0,0,32,,,,
+339,82,0,1,0,0,293,17,41,0,1,0,0,120,90,0,0,0,0,0,0,28,,,,
+340,61,0,0,0,0,252,7,4,0,1,0,0,140,40,0,0,1,0,0,0,74,,,,
+341,37,0,0,0,0,103,11,62,0,0,0,0,100,80,0,0,0,0,0,0,69,,,,
+342,18,0,0,0,0,49,4,27,0,1,0,0,150,60,1,0,1,0,0,1,179,,,,
+343,17,0,0,0,0,181,4,48,0,1,0,1,140,70,0,0,0,0,0,0,35,,,,
+344,49,0,1,0,0,242,0,67,0,0,1,1,100,90,0,0,0,0,0,0,127,,,,
+345,53,0,0,0,0,25,16,91,0,1,0,0,120,80,0,0,0,0,0,0,89,,,,
+346,7,0,1,0,0,32,0,41,0,1,0,0,150,80,1,1,1,0,0,0,30,,,,
+347,80,0,0,0,0,297,8,123,0,0,0,0,120,70,0,0,1,0,0,0,46,,,,
+348,48,0,0,0,0,247,17,129,0,1,0,0,150,70,0,0,0,0,0,1,193,,,,
+349,27,0,0,0,0,49,14,120,0,1,0,0,120,60,0,0,1,0,0,0,115,,,,
+350,21,0,1,0,0,264,4,27,0,1,0,0,150,70,0,0,0,0,0,0,101,,,,
+351,78,0,1,0,0,3,6,129,1,0,1,0,150,90,0,0,0,0,0,0,135,,,,
+352,1,0,0,0,1,189,14,55,0,1,0,0,130,70,0,0,0,0,0,0,1,,,,
+353,30,0,0,0,1,27,15,62,0,1,0,0,150,70,0,0,0,0,0,1,67,,,,
+354,35,0,0,0,0,88,11,106,0,0,0,0,140,60,0,0,0,0,0,0,83,,,,
+355,22,0,1,0,0,63,10,36,0,1,0,0,130,80,0,0,0,0,0,0,176,,,,
+356,9,0,0,0,0,182,16,47,0,0,0,0,110,90,0,0,1,0,0,1,60,,,,
+357,19,0,0,1,0,149,2,63,0,0,0,0,150,90,0,0,0,0,1,0,108,,,,
+358,84,0,0,0,0,151,16,39,1,1,1,1,120,90,0,0,0,0,1,1,40,,,,
+359,72,0,1,0,0,177,12,49,0,1,1,1,130,70,0,0,1,0,0,0,132,,,,
+360,54,0,1,0,0,109,12,122,0,1,0,0,90,80,0,0,0,1,0,1,137,,,,
+361,64,0,1,0,0,186,10,46,1,0,1,1,130,70,0,0,0,0,0,0,105,,,,
+362,69,0,0,0,0,76,4,67,0,1,0,0,100,60,0,0,0,0,0,1,121,,,,
+363,6,0,1,0,0,243,1,77,0,1,0,0,150,70,0,0,1,0,0,0,59,,,,
+364,7,0,1,0,0,2,5,81,0,0,0,0,120,80,0,0,1,0,0,0,175,,,,
+365,52,0,1,0,0,153,1,15,0,1,1,0,120,90,0,0,0,1,0,1,119,,,,
+366,16,0,0,0,0,302,11,29,0,0,0,0,110,40,0,0,0,0,1,0,107,,,,
+367,20,1,1,1,1,237,14,113,0,0,0,0,120,60,0,0,0,0,0,1,106,,,,
+368,27,0,0,0,0,239,4,10,0,1,0,0,110,90,0,0,0,0,0,0,131,,,,
+369,55,0,1,0,0,128,7,64,1,1,0,1,120,70,0,0,0,0,0,0,29,,,,
+370,9,0,1,0,0,4,13,2,0,0,0,1,130,70,0,0,1,0,0,0,9,,,,
+371,54,0,1,1,0,203,12,91,0,0,0,0,130,80,0,0,0,0,0,0,66,,,,
+372,86,0,1,0,0,157,13,68,1,0,1,0,120,70,0,0,0,0,0,0,127,,,,
+373,66,0,0,0,0,157,1,23,0,1,0,0,120,90,0,0,0,0,0,1,118,,,,
+374,42,0,1,0,0,235,10,50,0,1,0,1,130,90,0,0,0,0,0,0,122,,,,
+375,71,0,1,0,0,310,13,88,0,0,0,0,100,70,0,0,0,0,0,0,143,,,,
+376,53,1,0,1,0,39,7,11,0,1,0,0,120,70,0,0,0,0,0,0,166,,,,
+377,71,0,1,0,0,285,5,30,1,1,0,1,120,90,0,0,0,1,0,0,92,,,,
+378,22,0,0,1,0,138,13,99,0,1,0,0,130,80,0,1,1,0,0,0,130,,,,
+379,11,0,0,1,0,199,11,40,0,1,0,0,130,70,1,1,1,0,0,0,190,,,,
+380,29,0,1,0,0,224,0,67,0,1,0,0,110,90,0,0,1,0,0,0,183,,,,
+381,42,0,0,1,0,138,9,28,0,0,0,0,130,70,0,0,0,0,0,0,191,,,,
+382,70,0,1,0,0,100,12,97,0,0,0,0,140,50,0,0,0,0,0,0,126,,,,
+383,51,0,1,0,0,183,11,20,1,0,1,0,130,90,0,0,0,0,0,0,163,,,,
+384,29,0,0,0,0,221,3,93,0,1,0,0,130,90,0,0,1,0,0,0,81,,,,
+385,76,0,0,1,0,119,10,28,0,1,0,0,150,90,0,0,0,0,0,0,153,,,,
+386,73,0,1,0,0,34,12,76,0,0,0,0,130,40,0,0,0,0,0,0,83,,,,
+387,0,0,1,0,0,120,2,58,0,0,0,0,150,70,0,0,0,0,0,0,88,,,,
+388,72,0,0,0,0,301,5,80,0,1,0,0,120,80,0,0,1,0,0,1,177,,,,
+389,41,0,1,0,0,298,15,44,0,0,0,0,120,70,0,0,0,0,0,0,79,,,,
+390,61,0,1,0,0,54,13,103,1,1,1,1,120,90,0,0,0,0,0,0,37,,,,
+391,50,0,0,0,0,98,5,9,0,1,0,0,140,40,0,0,0,0,0,0,168,,,,
+392,76,0,1,0,0,72,16,113,0,1,0,0,140,70,0,0,1,0,0,0,125,,,,
+393,10,0,1,0,0,122,0,35,0,1,0,0,120,70,0,0,0,0,0,0,9,,,,
+394,2,0,1,0,0,113,10,109,0,0,0,0,120,90,0,0,0,0,0,0,94,,,,
+395,70,0,1,1,0,38,15,1,1,1,0,1,120,70,0,0,0,0,0,0,73,,,,
+396,79,0,1,0,1,178,10,118,0,1,0,0,110,80,0,0,0,0,0,0,182,,,,
+397,60,0,1,0,0,194,11,69,0,0,0,0,150,60,0,0,1,1,0,1,141,,,,
+398,74,0,0,0,0,211,10,36,0,1,0,1,150,40,0,0,1,0,0,0,20,,,,
+399,56,0,1,0,0,99,1,76,0,1,0,0,100,90,0,0,0,0,0,1,160,,,,
+400,62,0,1,0,0,276,1,108,0,0,0,0,140,80,0,0,0,0,0,0,67,,,,
+401,35,0,1,0,0,35,4,111,1,0,0,1,130,70,0,0,0,0,0,0,32,,,,
+402,63,0,0,0,0,83,3,13,0,0,0,0,140,70,0,0,0,0,0,1,139,,,,
+403,44,0,0,0,0,258,4,59,0,0,0,1,140,90,0,0,0,0,0,0,76,,,,
+404,74,0,1,1,0,290,17,98,0,0,0,0,150,80,0,0,0,0,0,1,89,,,,
+405,8,0,1,0,0,224,8,56,0,1,0,0,130,50,0,0,0,0,0,0,150,,,,
+406,70,0,1,0,0,163,9,16,1,0,0,0,120,90,0,0,0,0,0,0,10,,,,
+407,58,0,1,0,0,105,3,61,1,1,1,0,120,90,0,0,0,0,0,1,138,,,,
+408,53,0,1,0,0,15,7,23,0,0,0,0,150,50,0,0,0,0,0,0,26,,,,
+409,14,0,0,0,0,160,10,97,0,1,0,0,120,40,0,0,0,0,0,0,50,,,,
+410,52,0,1,0,1,198,17,37,0,1,0,0,100,40,0,1,0,0,0,0,111,,,,
+411,55,0,0,0,0,190,16,59,0,0,0,0,100,90,0,0,1,0,0,0,186,,,,
+412,84,0,0,0,0,51,15,115,0,0,0,0,140,80,0,0,0,0,0,0,155,,,,
+413,18,0,1,0,0,177,8,49,1,0,1,0,120,90,0,0,0,0,0,0,19,,,,
+414,8,0,0,0,0,96,17,17,0,1,0,0,120,80,0,0,0,0,0,1,93,,,,
+415,20,0,0,0,0,148,2,99,0,1,0,0,100,40,0,0,1,0,0,0,179,,,,
+416,68,1,0,1,0,229,15,94,0,1,0,0,100,40,0,0,0,0,0,0,43,,,,
+417,69,0,1,0,1,62,6,72,0,1,0,0,140,80,0,0,0,0,0,0,117,,,,
+418,0,0,0,0,0,120,11,126,0,1,0,0,100,90,0,0,0,0,0,0,119,,,,
+419,19,0,1,0,0,252,14,18,0,0,0,0,120,80,0,0,0,0,0,0,53,,,,
+420,61,0,1,0,0,237,12,83,0,0,0,0,120,70,0,0,0,0,0,0,129,,,,
+421,1,0,1,0,0,30,4,44,0,0,0,0,120,60,0,0,0,0,0,0,76,,,,
+422,13,0,1,0,0,287,13,12,0,0,0,0,130,70,0,0,0,0,0,0,28,,,,
+423,81,0,0,0,0,306,7,77,0,1,0,0,130,80,0,0,0,0,0,0,148,,,,
+424,84,0,0,0,1,103,5,120,0,0,0,0,130,70,0,0,0,0,1,0,91,,,,
+425,0,0,1,0,1,134,8,95,0,1,0,0,140,90,0,0,0,0,0,0,6,,,,
+426,38,0,0,0,0,212,6,122,0,1,0,0,140,60,0,0,1,0,0,0,11,,,,
+427,64,0,0,0,0,159,13,39,0,0,0,0,120,90,0,0,0,0,0,1,168,,,,
+428,13,0,1,0,0,305,13,91,0,0,0,0,150,50,0,0,1,0,0,0,44,,,,
+429,47,0,1,0,0,83,1,27,0,0,0,0,140,50,0,0,0,0,0,1,21,,,,
+430,32,0,0,0,0,311,3,31,0,1,0,0,120,90,0,0,0,0,0,1,134,,,,
+431,41,0,0,0,0,248,12,5,0,1,0,0,120,80,0,0,1,0,0,0,168,,,,
+432,13,0,1,0,0,36,16,74,0,0,0,0,140,70,0,0,0,0,0,0,155,,,,
+433,77,0,1,0,1,174,3,10,0,0,0,1,80,60,1,0,1,0,0,0,17,,,,
+434,2,1,0,1,1,239,15,114,0,1,0,0,140,90,0,0,0,0,0,0,149,,,,
+435,72,0,1,0,0,47,6,51,0,0,0,0,130,90,0,0,0,0,0,0,65,,,,
+436,14,0,0,0,0,16,1,77,0,0,0,0,150,90,0,0,0,0,0,1,18,,,,
+437,27,0,0,0,0,260,10,43,0,0,0,0,120,70,0,0,0,0,1,0,142,,,,
+438,57,0,1,0,0,5,17,1,0,0,0,0,120,90,1,1,1,0,0,1,57,,,,
+439,83,0,0,0,1,316,0,112,0,1,0,0,120,90,0,0,0,1,0,1,48,,,,
+440,19,0,1,0,0,160,0,81,0,0,0,0,150,60,0,0,0,0,1,0,56,,,,
+441,44,0,0,0,0,7,12,7,0,0,0,1,150,40,0,0,0,0,0,0,73,,,,
+442,23,0,1,0,0,203,12,53,0,1,0,0,140,70,0,0,0,0,0,0,181,,,,
+443,34,0,1,0,0,89,11,125,0,0,0,0,140,80,0,0,1,0,0,0,65,,,,
+444,10,0,0,0,0,195,15,99,0,0,0,1,100,90,0,0,0,0,0,0,99,,,,
+445,63,0,0,0,0,265,4,101,0,1,0,0,150,60,0,0,0,0,0,0,3,,,,
+446,82,0,1,0,0,128,14,56,0,1,0,1,130,80,0,0,1,0,0,0,141,,,,
+447,5,0,1,0,0,124,16,26,0,1,0,0,100,70,0,0,0,0,0,1,17,,,,
+448,39,0,0,0,0,124,9,8,0,0,0,0,120,50,0,1,1,0,0,1,158,,,,
+449,86,0,1,0,0,48,16,62,0,0,0,0,120,60,0,0,0,0,0,1,110,,,,
+450,29,0,1,0,0,188,9,72,0,0,0,0,150,60,0,0,0,0,0,1,25,,,,
+451,45,0,0,1,0,29,3,1,1,0,1,1,170,90,0,0,1,0,0,0,145,,,,
+452,0,0,1,0,0,271,5,124,0,1,0,0,140,50,0,0,0,0,0,1,34,,,,
+453,47,1,0,1,0,6,5,1,0,0,0,0,100,70,0,0,0,0,0,0,10,,,,
+454,85,1,0,1,1,68,16,71,1,0,0,1,120,70,0,1,0,0,0,1,112,,,,
+455,85,1,1,1,1,129,6,69,0,1,0,0,110,70,0,0,0,0,1,1,98,,,,
+456,79,0,0,0,1,144,3,28,0,0,0,0,150,70,1,0,1,0,0,0,188,,,,
+457,4,0,0,1,0,142,15,116,0,1,0,0,140,90,0,0,1,0,0,0,40,,,,
+458,39,0,0,0,0,202,2,80,0,0,0,0,150,70,0,0,1,0,0,1,86,,,,
+459,82,0,1,0,0,209,9,43,1,0,1,0,130,90,0,0,0,1,0,0,148,,,,
+460,79,0,0,0,0,123,8,44,0,0,0,0,110,80,0,0,0,0,0,0,186,,,,
+461,53,0,1,0,0,28,7,100,0,1,0,0,150,80,0,0,1,1,0,1,136,,,,
+462,51,0,0,0,0,46,2,87,0,0,0,0,100,50,0,0,0,0,0,0,24,,,,
+463,32,0,0,0,0,246,4,66,0,1,0,0,110,70,0,0,0,0,0,1,22,,,,
+464,76,0,0,0,0,131,15,22,0,0,0,0,80,90,0,0,0,0,0,1,182,,,,
+465,11,0,0,0,0,147,14,90,0,1,0,0,110,60,0,0,1,0,0,0,54,,,,
+466,43,0,0,0,0,16,2,90,1,0,0,1,120,90,0,0,0,1,0,1,2,,,,
+467,3,0,1,0,0,311,5,12,0,0,0,0,140,60,0,0,0,0,0,0,173,,,,
+468,27,0,1,0,0,260,3,31,0,0,1,0,120,90,0,0,0,0,0,0,33,,,,
+469,78,0,1,1,0,262,5,122,0,0,0,0,150,70,0,0,0,0,0,0,164,,,,
+470,84,0,0,0,0,266,9,55,0,1,0,0,110,90,0,0,0,0,0,1,68,,,,
+471,58,0,1,0,0,180,9,4,0,0,0,0,100,90,0,0,0,0,0,1,48,,,,
+472,62,0,0,0,0,25,16,85,0,1,0,0,140,80,0,0,0,0,0,1,183,,,,
+473,76,0,0,0,1,155,0,16,1,0,0,1,130,90,0,0,0,0,0,0,5,,,,
+474,47,0,1,0,1,271,7,45,0,1,0,0,120,90,0,0,0,0,0,0,43,,,,
+475,0,0,0,0,0,238,16,56,0,0,0,0,140,70,0,0,0,0,0,1,106,,,,
+476,51,0,1,0,0,300,8,90,0,0,0,0,90,80,0,0,0,0,0,1,21,,,,
+477,51,0,0,0,0,151,16,104,0,1,0,0,150,50,0,0,0,0,0,1,39,,,,
+478,49,0,1,0,0,303,16,29,0,0,0,0,140,60,0,0,0,0,0,0,118,,,,
+479,81,0,0,0,0,47,4,107,1,1,1,1,120,90,0,0,0,0,0,0,24,,,,
+480,45,0,1,0,0,185,9,118,0,1,0,0,150,90,0,0,0,0,1,0,98,,,,
+481,62,0,1,0,1,92,10,82,0,0,0,0,130,70,0,0,0,0,0,1,146,,,,
+482,73,1,1,1,0,279,4,118,0,1,0,0,150,60,0,0,1,0,0,0,151,,,,
+483,3,0,0,0,0,73,1,80,0,0,0,0,130,80,0,0,0,1,0,1,103,,,,
+484,55,0,1,0,1,21,9,75,1,1,0,1,120,70,0,0,0,0,0,0,194,,,,
+485,24,0,0,0,0,64,4,34,0,1,0,0,110,80,0,0,0,0,0,1,82,,,,
+486,31,0,1,0,1,303,10,14,0,0,0,0,100,60,0,0,0,0,0,0,146,,,,
+487,87,0,1,0,0,10,6,7,0,0,0,1,110,60,0,0,0,0,0,0,84,,,,
+488,25,0,1,0,1,79,9,89,0,1,0,0,110,90,0,0,0,0,0,0,81,,,,
+489,86,0,1,0,0,187,9,53,0,1,0,0,110,80,0,0,0,0,0,0,121,,,,
+490,25,0,0,0,0,172,16,61,0,1,0,0,100,70,0,0,0,0,0,0,192,,,,
+491,57,0,1,0,0,3,7,68,0,1,0,0,130,90,0,0,0,0,0,0,15,,,,
+492,26,0,1,0,0,107,0,115,0,0,0,0,140,70,0,0,0,1,0,1,172,,,,
+493,73,0,1,0,0,106,14,42,0,1,0,0,80,50,0,0,0,0,0,1,102,,,,
+494,61,1,1,1,1,255,2,97,0,0,0,0,100,40,0,0,1,0,0,1,144,,,,
+495,81,0,1,0,0,319,12,26,0,1,0,0,120,90,0,0,1,0,0,0,165,,,,
+496,75,1,0,1,1,102,7,3,0,0,0,0,80,50,0,0,0,0,0,0,147,,,,
+497,40,0,0,0,0,121,5,93,0,0,0,0,100,80,0,0,0,1,0,1,140,,,,
+498,18,0,0,0,1,53,10,74,0,0,0,0,130,90,0,1,1,0,0,1,151,,,,
+499,2,0,0,0,0,64,2,96,0,1,0,0,140,80,0,0,0,0,0,0,53,,,,

README.md

@@ -1,13 +1,52 @@
 ---
-title: XLabel
+title: "XLabel: eXplainable Labeling Assistant"
 emoji: 💻
-colorFrom: indigo
+colorFrom: pink
 colorTo: gray
 sdk: streamlit
-sdk_version: 1.17.0
+sdk_version: 1.15.2
 app_file: app.py
-pinned: false
+pinned: true
 license: apache-2.0
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# XLabel: e**X**plainable **Label**ing Assistant
+
+XLabel is an open-source [Streamlit](https://streamlit.io/) app that takes an explainable machine learning approach to visual-interactive data labeling.
+
+This is the official code of the following paper:
+[An Explainable Machine Learning Approach to Visual-Interactive Labeling: A Case Study on Non-communicable Disease Data](https://arxiv.org/abs/2209.12778)
+Donlapark Ponnoprat, Parichart Pattarapanitchai, Phimphaka Taninpong, Suthep Suantai
+
+## News (01/01/2023)
+* Use tabs instead of radio buttons for multiple labels.
+* The app now requires `streamlit>=1.16.0` for the tabs and `interpret>=0.3.0` for handling missing data.
+
+## Features
+XLabel can:
+* Predict the most probable labels using Explainable Boosting Machine (EBM).
+* Show the contributions of each feature towards the predicted labels.
+* Provide an option to write the labels directly into the data file (use `XLabel.py`) or save them in a separate file (use `XLabelDL.py`)
+* Support data with multiple labels and multiple classes.
+* Support data with missing values ([thanks to EBM](https://github.com/interpretml/interpret/issues/18)) and/or non-numeric categorical features.
+
+## Usage
+Before using XLabel, the data file must follow the following tabular convention:
+* The file must be in either CSV or Excel format.
+* The first row of the file must be the names of the columns.
+* The first column must contain a unique identifier (id) for each row.
+* The label columns must appear last.
+In addition, a few instances must have already been labeled, with each class appearing at least once (For example, if a label has five possible classes, then the required minimum number of labeled instances is 5).
+
+With your data file satisfying these conditions, you can now start data labeling with XLabel!
+1. Copy `XLabel.py` to the directory that contains the data file and run the `streamlit` command:
+    ```
+    streamlit run XLabel.py
+    ```
+    * By design, `XLabel.py` will write the labeled data to the original data file. If instead you would like to download the labeled data as a separate file, use `XLabelDL.py` instead.
+    * You can assign a specific list of input features for each label by editing `configs.json` and copying it along with `XLabel.py`. There are also other sidebar options that you can play around as well. Here is an example ofr [`configs.json`](configs.json).
+2. Upload a data file (only on the first run), select the options on the sidebar, and then click "**Sample**". The samples with lowest predictive confidences will be shown first in the main screen.
+3. Check the suggested labels; you can keep the correct ones and change the wrong ones.
+4. Click the "**Submit Labels**" button at the bottom of the page to save the labels. 
+    * If you are using `XLabel.py`, the labels will be saved directly to the original data file.
+    * If you are using `XLabelDL.py`, you need to click the `Download labeled data` in the sidebar to download the labeled data as a new file.

app.py

@@ -0,0 +1,614 @@
+"""A Streamlit app designed to help with data labeling with explainable
+machine learning approach. It can handle data with many labels and many
+classes. For each label, an Explainable Boosting Machine model is
+trained on the labeled data, then it makes class predictions and
+provides per-instance local explanations, which are then used to make
+heatmaps, displayed in the main screen.
+Model page: https://huggingface.co/Donlapark/XLabel
+"""
+import json
+import math
+import os
+import pickle as pkle
+
+from interpret.glassbox.ebm.ebm import ExplainableBoostingClassifier
+
+import numpy as np
+import pandas as pd
+
+import altair as alt
+import streamlit as st
+from streamlit import session_state as _state
+import streamlit.components.v1 as components
+
+_PERSIST_STATE_KEY = f"{__name__}_PERSIST"
+_CONFIGS_FILE = "configs.json"
+_MODEL = "_saved_models.pickle"
+_NUM_FEAT_PER_ROW = 11
+
+st.set_page_config(layout="wide")
+
+
+def main():
+    """The main Streamlit app."""
+    if "configs" not in _state:
+        try:
+            with open(_CONFIGS_FILE, "r") as _file:
+                _state["configs"] = json.load(_file)
+        except FileNotFoundError:
+            create_config_file()
+
+        _state["loaded_new_file"] = True
+
+    st.sidebar.write("Current database: " + _state.configs["db_filename"])
+
+    st.sidebar.file_uploader("Upload a CSV or Excel file",
+                             type=["csv", "xlsx", "xls"],
+                             key="uploaded_files",
+                             accept_multiple_files=False,
+                             on_change=update_file)
+
+    with st.sidebar.form("sidebar"):
+        st.slider("Number of labels",
+                  min_value=1,
+                  max_value=20,
+                  value=_state.configs["sidebar"]["num_labels"],
+                  step=1,
+                  key="num_labels")
+
+        st.selectbox("Include data with label/prediction mismatches?",
+                     ("Yes", "No"),
+                     key="relabel",
+                     index=("Yes",
+                            "No").index(_state.configs["sidebar"]["relabel"]))
+
+        st.selectbox("Sampling mode",
+                     ("Fixed sample size", "Confidence threshold"),
+                     key="mode",
+                     index=("Fixed sample size", "Confidence threshold").index(
+                         _state.configs["sidebar"]["mode"]))
+
+        st.slider("Sample size (for \"Fixed sample size\" mode)",
+                  min_value=1,
+                  max_value=500,
+                  value=_state.configs["sidebar"]["n_samples"],
+                  step=1,
+                  key="n_samples")
+
+        st.slider("Threshold (for \"Confidence threshold\" mode)",
+                  min_value=0.00,
+                  max_value=1.00,
+                  value=_state.configs["sidebar"]["threshold"],
+                  step=0.01,
+                  format="%.2f",
+                  key="threshold")
+
+        form_cols = st.columns((2, 2, 2))
+        form_cols[1].form_submit_button("Sample", on_click=sample_and_predict)
+
+    if "pages" in _state:
+        page_list = list(_state.pages)
+        tabs = st.tabs(page_list)
+        for i, tab in enumerate(tabs):
+            with tab:
+                display_main_screen(page_list[i])
+
+    filename = _state.configs["db_filename"]
+    file_pre, file_ext = os.path.splitext(filename)
+    if "database" in _state:
+        data, mime = convert_to_downloadable(_state.database, file_ext)
+
+        st.sidebar.download_button(label="Download labeled data",
+                                   data=data,
+                                   file_name=filename,
+                                   mime=mime)
+
+
+def update_file():
+    """Update the state parameters after a file has been uploaded"""
+    if _state.uploaded_files is not None:
+        _state.configs["db_filename"] = _state.uploaded_files.name
+        _state.loaded_new_file = True
+
+
+def init_state_params():
+    """Initialize all state parameters.
+
+    This function will be called by sample_and_predict() when
+    _state.pages has not been initialized.
+
+    State parameters:
+        database: The pandas dataframe of the database.
+        configs: The saved configs of sidebar widgets.
+        pages: The list of all labels.
+        classes: The dict of all classes for each label.
+        class_to_num: The encoding dict of classes into integers.
+        num_to_class: The decoding dict of integers into classes.
+        class_to_num: The dict that maps each class to the
+          corresponding number.
+        previous_: The index of the previous page.
+        next_: The index of the next page.
+        next_clicked: The index of the current page.
+        local_results: A dict of outputs of EBM
+          used to write predictions and plot heatmaps on screen.
+        models: A dict of EBM models to predict the labels.
+        models_params: A dict of models' attributes, which will be
+          saved as a pickle file.
+        predictions: A pandas dataframe; each column contains EBM's
+          predictions of each label.
+        unlabeled_index: A pandas index of unlabeled rows. When new
+          labels are added to the database, compute_unlabeled_index()
+          needs to be called to track the changes.
+    """
+    if _state.uploaded_files is not None:
+        data_file = _state.uploaded_files
+        _state.configs["db_filename"] = data_file.name
+    else:
+        data_file = _state.configs["db_filename"]
+
+    filename = _state.configs["db_filename"]
+    if filename == "None":
+        return
+
+    file_pre, file_ext = os.path.splitext(filename)
+    if file_ext == ".csv":
+        _state.database = pd.read_csv(data_file, index_col=0)
+    elif (file_ext == ".xlsx") or (file_ext == ".xls"):
+        _state.database = pd.read_excel(data_file, index_col=0)
+
+    create_pages()
+
+
+def create_pages():
+    """Add or change state parameters that are related to labeling pages
+
+    These parameters assign the labels to multiple pages, with
+    one label per page.
+    """
+    _state["pages"] = _state.database.columns[-_state.num_labels:]
+    _state["classes"] = {
+        label: sorted(list(_state.database[label].dropna().unique()))
+        for label in _state.pages
+    }
+    _state["num_to_class"] = {
+        label: dict(enumerate(_state.classes[label]))
+        for label in _state.pages
+    }
+    _state["class_to_num"] = {
+        label: {c: i
+                for i, c in enumerate(_state.classes[label])}
+        for label in _state.pages
+    }
+
+    _state.update({
+        'local_results': {}
+    })
+
+    _state["predictions"] = pd.DataFrame(index=_state.database.index,
+                                         columns=_state.pages)
+
+    file_pre, file_ext = os.path.splitext(_state.configs["db_filename"])
+    try:
+        with open(file_pre + str(_state.num_labels) + _MODEL, 'rb') as _file:
+            _state["models_params"] = pkle.load(_file)
+            _state["models"] = {}
+            for label in _state.pages:
+                _state.models[label] = ExplainableBoostingClassifier()
+                _state.models[label].__dict__.update(
+                    _state.models_params[label])
+    except FileNotFoundError:
+        _state["models"], _state["models_params"] = initialize_models()
+
+    compute_unlabeled_index()
+
+
+def initialize_models():
+    """initialize and train EBMs for all labels.
+
+    If a pickle file of EBM models (stored in _MODEL) is not found
+    in the directory, this function will be called by init_state_params()
+    to initialize the models.
+    """
+    models = {}
+    models_params = {}
+    for label in _state.pages:
+        y = _state.database[label].dropna().map(_state.class_to_num[label])
+        X = subset_features(_state.database, label)
+        X = X.loc[y.index, :]
+        models[label] = ExplainableBoostingClassifier().fit(X, y)
+        models_params[label] = models[label].__dict__
+
+    return models, models_params
+
+
+def subset_features(X, label):
+    """Returns a subset of features specified in state's input_features parameters
+    Args:
+        X: a Pandas DataFrame.
+        label: The column name of the labels.
+
+    Returns
+        A Pandas DataFrame consisting of a subset of features in X.
+    """
+    input_features = _state.configs["input_features"]
+    if label not in input_features.keys():
+        X = X.iloc[:, :-_state.num_labels]
+    else:
+        X = X.loc[:, input_features[label]]
+    return X
+
+
+def compute_unlabeled_index(new_labeled_index=None, label=None):
+    """Track the indices of unlabeled data after introducing new labels.
+
+    Args:
+        new_labeled_index: A pandas index of newly labeled data.
+        label: The column name of the new labels.
+    """
+    if new_labeled_index is not None:
+        _state.unlabeled_index[label] = _state.unlabeled_index[
+            label].difference(new_labeled_index)
+    else:
+        all_index = _state.database.index
+        _state.unlabeled_index = {
+            label: all_index[_state.database[label].isna()]
+            for label in _state.pages
+        }
+
+
+def create_config_file():
+    """Create a new config file"""
+    _state["configs"] = {
+        "db_filename": "None",
+        "sidebar": {
+            "num_labels": 1,
+            "relabel": "Yes",
+            "mode": "Fixed sample size",
+            "n_samples": 50,
+            "threshold": 0.95
+        },
+        "input_features": {}
+    }
+    with open(_CONFIGS_FILE, "w") as _file:
+        json.dump(_state.configs, _file, indent=4)
+
+
+@st.experimental_memo
+def convert_to_downloadable(data, file_type):
+    """Convert a dataframe to a downloadable format
+
+    Args:
+    data: A Pandas DataFrame.
+    file_type: A target format for the conversion: ".csv", ".xls" or ".xlsx".
+
+    Returns:
+    converted_data: Data converted to the specified format.
+    """
+    if file_type == ".csv":
+        converted_data = data.to_csv().encode('utf-8')
+        mime = "text/csv"
+    elif (file_type == ".xlsx") or (file_type == ".xls"):
+        converted_data = data.to_excel().encode('utf-8')
+        mime = "application/vnd.ms-excel"
+    else:
+        raise ValueError("file_type must be \"csv\" or \"excel\"")
+
+    return converted_data, mime
+
+
+def display_main_screen(label):
+    """Display predictions and heatmaps on the main screen.
+
+    This function is called after EBM has been trained on the labeled data.
+    The predictions and explanations (displayed as heatmaps) will be shown
+    on the main screen.
+
+    Args:
+        label: the column name of the predictions.
+    """
+    main_cols = st.columns((4, 4, 4))
+    if _state.unlabeled_index[label].empty:
+        main_cols[1].write("All " + label + " data are labeled.")
+    else:
+        with st.form(label + "Label form"):
+            if _state.local_results[label] == {}:
+                main_cols[1].write("""There are some unlabeled data left.  \n \
+                This means that the confidences of the remaining data are \
+                above the threshold. \n You can either let the model label \
+                these data automatically \n or change the sampling mode to \
+                \"Fixed sample size\".""")
+            else:
+                input_features = _state.configs["input_features"]
+                if label not in input_features.keys():
+                    num_features = _state.database.shape[1] - _state.num_labels
+                else:
+                    num_features = len(input_features[label])
+                num_heatmap_rows = math.ceil(num_features / _NUM_FEAT_PER_ROW)
+                for page in _state.local_results[label]:
+                    current_plot = plot_all_features(
+                        _state.local_results[label][page]['data'],
+                        title=str(page),
+                        height=50,
+                        num_rows=num_heatmap_rows)
+                    cols = st.columns((6, 1))
+                    #with cols[0]:
+                    #    if _state.text1 is not None:
+                    #        st.write(_state.data[_state.text1][page])
+                    #    if _state.text2 is not None:
+                    #        st.write(_state.data[_state.text2][page])
+
+                    cols[0].altair_chart(current_plot,
+                                         use_container_width=True)
+
+                    prediction = _state.local_results[label][page][
+                        'prediction']
+                    cols[1].radio("Label",
+                                  options=_state.classes[label],
+                                  key=label + str(page),
+                                  index=int(prediction))
+                    results = report_results(page, label)
+                    for result in results:
+                        cols[1].write(result)
+                    st.markdown("""---""")
+
+            label_from_cols = st.columns((4, 4, 4))
+
+            label_from_cols[1].radio("Automatically label the remaining data?",
+                                     ("Yes", "No"),
+                                     index=1,
+                                     key=label+"_auto")
+
+            label_from_cols[1].form_submit_button("Submit Labels",
+                                                  on_click=update_and_save,
+                                                  args=(label, ))
+
+
+@st.experimental_memo
+def plot_all_features(data, title, height, num_rows):
+    """Plot all rows of the heatmap of EBM's per-instance explanation.
+
+    Args:
+        data: Per-instance local explanations from EBM.
+        title: The plot's title.
+        height: The height of the plot.
+        num_rows: The number of rows of the heatmap.
+
+    Returns:
+        obj: An Altair plot object.
+    """
+    plot_list = [None] * num_rows
+    if num_rows == 1:
+        plot_list[0] = plot(data, title, height)
+    else:
+        plot_list[0] = plot(data.iloc[0:_NUM_FEAT_PER_ROW], title, height)
+        for i in range(1, num_rows - 1):
+            plot_list[i] = plot(
+                data.iloc[_NUM_FEAT_PER_ROW * i:_NUM_FEAT_PER_ROW * (i + 1)],
+                "", height)
+        plot_list[-1] = plot(data.iloc[_NUM_FEAT_PER_ROW * (num_rows - 1):],
+                             "", height)
+
+    obj = alt.vconcat(*plot_list).configure_axis(
+        labelFontSize=13, titleFontSize=16, labelAngle=0,
+        title=None).configure_title(fontSize=16)
+
+    return obj
+
+
+def plot(data, title, height):
+    """Plot each row of the heatmap of EBM's per-instance explanation.
+
+    Args:
+        data: Per-instance local explanations from EBM.
+        title: The plot's title.
+        height: The height of the plot.
+
+    Returns:
+        obj: An Altair plot object.
+    """
+    base = alt.Chart(data).encode(x=alt.X('features', sort=None))
+
+    heatmap = base.mark_rect().encode(color=alt.Color(
+        'scores:Q',
+        scale=alt.Scale(scheme='redblue', reverse=True, domain=[0, 1]),
+        legend=alt.Legend(direction='vertical')))
+
+    # Configure text
+    text = base.mark_text(baseline='middle', fontSize=14).encode(
+        text='values:N',
+        color=alt.condition(
+            (alt.datum.scores > 0.8) | (alt.datum.scores < 0.2),
+            alt.value('white'), alt.value('black')))
+
+    obj = (heatmap + text).properties(height=height, width=550, title=title)
+
+    return obj
+
+
+@st.experimental_memo
+def report_results(idx, col_name):
+    """Create a list that contains current label (if exists) and confidence score.
+
+    Args:
+        idx: A row's index in the database.
+        col_name: A column's name in the database.
+
+    Returns:
+        results: A list of current label (if exists) and confidence score.
+    """
+    results = []
+    current_label = _state.database[col_name][idx]
+    if not pd.isna(current_label):
+        results.append(f"Current label: {current_label}")
+
+    confidence = _state.local_results[col_name][idx]['confidence']
+    results.append(f"Confidence: {confidence:.2f}")
+
+    return results
+
+
+def sample_and_predict():
+    """Sample data and make a dict of predictions and explanations.
+
+    This function calls EBM to predict the labels and give per-instance
+    local explanations. This function calls generate_explanation() to store
+    the predictions and explanations in a dictionary.
+    """
+    st.experimental_memo.clear()
+
+    if _state.loaded_new_file:
+        init_state_params()
+        _state.loaded_new_file = False
+    else:
+        if "database" not in _state:
+            st.error("No database has been uploaded.")
+            return
+        if _state.configs["sidebar"]["num_labels"] != _state.num_labels:
+            create_pages()
+
+    _state.local_results = dict.fromkeys(_state.pages)
+
+    for label in _state.pages:
+        X = subset_features(_state.database, label)
+        if _state.relabel == "No":
+            X_unlabeled = X.loc[_state.unlabeled_index[label], :]
+        else:
+            X_unlabeled = X
+        _state.local_results[label] = {}
+
+        model = _state.models[label]
+        generate_explanation(X_unlabeled, label, model)
+
+    for k in _state.configs["sidebar"].keys():
+        _state.configs["sidebar"][k] = _state[k]
+    with open(_CONFIGS_FILE, "w") as _file:
+        json.dump(_state.configs, _file, indent=4)
+
+
+def update_and_save(label):
+    """Update the labels, then retrain and save the models.
+
+    Store the user's labels in the database, which is then saved to
+    a local disk. EBM is then retrained on the database with addition
+    labels, after which, a new list of predictions and explanations
+    will be shown on the main screen. This function calls
+    generate_explanation() to store the predictions and explanations
+    in a dictionary.
+
+    Args:
+        label: the column name of the label.
+    """
+    new_labeled_index = list(_state.local_results[label].keys())
+    _state.database.loc[new_labeled_index, label] = [
+        _state[label + str(ix)] for ix in new_labeled_index
+    ]
+    compute_unlabeled_index(new_labeled_index, label)
+
+    if _state[label + "_auto"] == "Yes":
+        unlabeled_idx = _state.unlabeled_index[label]
+        class_pred = _state.predictions.loc[unlabeled_idx, label]
+        _state.database.loc[unlabeled_idx, label] = class_pred
+        _state.unlabeled_index[label] = pd.Index([])
+        labeled_index = _state.database.index
+    else:
+        labeled_index = _state.database.index.difference(
+            _state.unlabeled_index[label])
+
+    X = subset_features(_state.database, label)
+    X_train = X.loc[labeled_index, :]
+    ytrain = _state.database.loc[labeled_index, label]
+    ebm = ExplainableBoostingClassifier()
+    ebm.fit(X_train, ytrain.map(_state.class_to_num[label]))
+    _state.models[label] = ebm
+    _state.models_params[label] = ebm.__dict__
+
+    filename = _state.configs["db_filename"]
+    file_pre, file_ext = os.path.splitext(filename)
+    with open(file_pre + str(_state.num_labels) + _MODEL, 'wb') as _file:
+        pkle.dump(_state.models_params, _file, protocol=pkle.HIGHEST_PROTOCOL)
+
+    _state.local_results[label] = {}
+    if _state[label + "_auto"] == "No":
+        X = X.loc[_state.unlabeled_index[label], :]
+        generate_explanation(X, label, ebm)
+
+
+def generate_explanation(X, label, model):
+    """Create a dict of predictions and explanations of a sample.
+
+    Make label predictions and per-instance local explanations,
+    which are then stored as a dict in _state.local_results.
+
+    Args:
+        X: A set of unlabeled data.
+        label: The column name of a label.
+        model: A model to predict labels and provide explanations.
+    """
+    n_samples = X.shape[0]
+    n_features = X.shape[1]
+
+    localx = model.explain_local(X)._internal_obj['specific']
+    ypred = np.array([
+        _state.num_to_class[label][localx[j]['perf']['predicted']]
+        for j in range(n_samples)
+    ])
+    _state.predictions.loc[X.index, label] = ypred
+    y = _state.database.loc[X.index, label]
+
+    p = np.array(
+        [localx[j]['perf']['predicted_score'] for j in range(n_samples)])
+    scores = np.minimum(p, (pd.isnull(y) | (ypred == y)))
+
+    if _state.mode == "Confidence threshold":
+        top_ind = np.where(scores <= _state.threshold)[0]
+    else:
+        n_samples = np.minimum(_state.n_samples, scores.shape[0] - 1)
+        top_ind = np.argpartition(scores, n_samples)[:n_samples]
+
+    X_ = X.iloc[top_ind, :].copy()
+    ypred = ypred[top_ind]
+
+    id_idx_pair = dict(zip(X_.index, top_ind))
+
+    try:
+        data_by_class = [X_[ypred == c] for c in _state.classes[label]]
+    except KeyError:
+        return
+
+    feature_names = X.columns
+
+    for sgn_data in data_by_class:
+        current_dict = _state.local_results[label]
+        for j in sgn_data.index:
+            localxi = localx[id_idx_pair[j]]
+
+            if len(_state.classes[label]) == 2:
+                feature_contrib = localxi['scores'][:n_features]
+            else:
+                feature_contrib = [
+                    localxi['scores'][k][localxi['perf']['predicted']]
+                    for k in range(n_features)
+                ]
+            heatmap_data = pd.DataFrame({
+                'features':
+                feature_names,
+                'values':
+                localxi['values'][:n_features],
+                'scores':
+                1 / (1 + 1 / np.exp(feature_contrib))
+            })
+            heatmap_data = heatmap_data.astype({
+                'features': str,
+                'values': str,
+                'scores': float
+            })
+            current_dict[j] = {
+                'actual': localxi['perf']['actual'],
+                'prediction': localxi['perf']['predicted'],
+                'confidence': localxi['perf']['predicted_score'],
+                'data': heatmap_data
+            }
+
+
+if __name__ == "__main__":
+    main()

configs.json

@@ -0,0 +1,37 @@
+{
+    "db_filename": "NCD_synthetic.csv",
+    "sidebar": {
+        "num_labels": 4,
+        "relabel": "Yes",
+        "mode": "Fixed sample size",
+        "n_samples": 50,
+        "threshold": 0.95
+    },
+    "input_features": {
+        "DM_label": [
+            "DM_key",
+            "DM_ICD10",
+            "DM_drugs",
+            "Glucose",
+            "HbA1c"
+        ],
+        "HTN_label": [
+            "HTN_key",
+            "HTN_ICD10",
+            "HTN_drugs",
+            "sbp1",
+            "dbp1"
+        ],
+        "CKD_label": [
+            "CKD_key",
+            "CKD_ICD10",
+            "eGFR"
+        ],
+        "DLP_label": [
+            "DLP_key",
+            "DLP_ICD10",
+            "AGE",
+            "LDL-c"
+        ]
+    }
+}

requirements.txt

@@ -0,0 +1 @@
+interpret==0.3.2