[o] GLIM 3.77 update 1 (copyright)1985 Royal Statistical Society, London [o] [i] ? $units 88 $data id nox c e $dinput 11 80 $ [i] File name? ethanol [i] 1 3.741 12.0 0.907 [i] 2 2.295 12.0 0.761 [i] 3 1.498 12.0 1.108 [i] 4 2.881 12.0 1.016 [i] 5 0.760 12.0 1.189 [i] 6 3.120 9.0 1.001 [i] 7 0.638 9.0 1.231 [i] 8 1.170 9.0 1.123 [i] 9 2.358 12.0 1.042 [i] 10 0.606 12.0 1.215 [i] 11 3.669 12.0 0.930 [i] 12 1.000 12.0 1.152 [i] 13 0.981 15.0 1.138 [i] 14 1.192 18.0 0.601 [i] 15 0.926 6.0 0.696 [i] 16 1.590 12.0 0.686 [i] 17 1.806 12.0 1.072 [i] 18 1.962 15.0 1.074 [i] 19 4.028 15.0 0.934 [i] 20 3.148 9.0 0.808 [i] 21 1.836 9.0 1.071 [i] 22 2.845 6.0 1.009 [i] 23 1.013 6.0 1.142 [i] 24 0.414 18.0 1.229 [i] 25 0.812 18.0 1.175 [i] 26 0.374 15.0 0.568 [i] 27 3.623 15.0 0.977 [i] 28 1.869 6.0 0.767 [i] 29 2.836 6.0 1.006 [i] 30 3.567 9.0 0.893 [i] 31 0.866 15.0 1.152 [i] 32 1.369 15.0 0.693 [i] 33 0.542 15.0 1.232 [i] 34 2.739 15.0 1.036 [i] 35 1.200 15.0 1.125 [i] 36 1.719 9.0 1.081 [i] 37 3.423 9.0 0.868 [i] 38 1.634 6.0 0.762 [i] 39 1.021 6.0 1.144 [i] 40 2.157 6.0 1.045 [i] 41 3.361 18.0 0.797 [i] 42 1.390 18.0 1.115 [i] 43 1.947 18.0 1.070 [i] 44 0.962 18.0 1.219 [i] 45 0.571 9.0 0.637 [i] 46 2.219 9.0 0.733 [i] 47 1.419 9.0 0.715 [i] 48 3.519 9.0 0.872 [i] 49 1.732 6.0 0.765 [i] 50 3.206 6.0 0.878 [i] 51 2.471 6.0 0.811 [i] 52 1.777 15.0 0.676 [i] 53 2.571 18.0 1.045 [i] 54 3.952 18.0 0.968 [i] 55 3.931 15.0 0.846 [i] 56 1.587 15.0 0.684 [i] 57 1.397 6.0 0.729 [i] 58 3.536 6.0 0.911 [i] 59 2.202 6.0 0.808 [i] 60 0.756 6.0 1.168 [i] 61 1.620 6.0 0.749 [i] 62 3.656 6.0 0.892 [i] 63 2.964 6.0 1.002 [i] 64 3.760 18.0 0.812 [i] 65 0.672 18.0 1.230 [i] 66 3.677 18.0 0.804 [i] 67 3.517 12.0 0.813 [i] 68 3.290 12.0 1.002 [i] 69 1.139 9.0 0.696 [i] 70 0.727 9.0 1.199 [i] 71 2.581 9.0 1.030 [i] 72 0.923 15.0 0.602 [i] 73 1.527 15.0 0.694 [i] 74 3.388 15.0 0.816 [i] 75 2.085 15.0 1.037 [i] 76 0.966 15.0 1.181 [i] 77 3.488 6.0 0.899 [i] 78 0.754 6.0 1.227 [i] 79 0.797 9.0 1.180 [i] 80 2.064 6.0 0.795 [i] 81 3.732 18.0 0.990 [i] 82 0.586 18.0 1.201 [i] 83 0.561 6.0 0.629 [i] 84 0.563 9.0 0.608 [i] 85 0.678 12.0 0.584 [i] 86 0.370 15.0 0.562 [i] 87 0.530 18.0 0.535 [i] 88 1.900 18.0 0.655 [i] ? $plot nox c $ [o] 4.200 | [o] 4.000 | 2 N [o] 3.800 | N 2 [o] 3.600 | 2 2 2 N N [o] 3.400 | N N N N [o] 3.200 | N 2 N [o] 3.000 | N [o] 2.800 | 2 N N [o] 2.600 | N N [o] 2.400 | N N [o] 2.200 | 2 N N [o] 2.000 | N 2 2 [o] 1.800 | 2 2 N N [o] 1.600 | 2 N 2 [o] 1.400 | N N N N N [o] 1.200 | 2 N N [o] 1.000 | 3 N 3 N [o] 0.800 | 2 2 N N N [o] 0.600 | N 3 2 N 3 [o] 0.400 | 2 N [o] 0.200 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 5.00 7.50 10.00 12.50 15.00 17.50 20.00 [i] ? $plot nox e $ [o] 4.200 | [o] 4.000 | N N N [o] 3.800 | N N N [o] 3.600 | NN N 2NN N [o] 3.400 | NN N N [o] 3.200 | N N 2 [o] 3.000 | N [o] 2.800 | 2NN [o] 2.600 | NN [o] 2.400 | N N [o] 2.200 | N N N N [o] 2.000 | N N N 2 [o] 1.800 | N 2 2N [o] 1.600 | 3 NN [o] 1.400 | N NN NN [o] 1.200 | N N 2 [o] 1.000 | N N 22 N N [o] 0.800 | N22N N [o] 0.600 | N NNNN NN3 [o] 0.400 | NN N [o] 0.200 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.480 0.640 0.800 0.960 1.120 1.280 1.440 [i] ? $cal lnox=%log(nox) $plot lnox e $ [o] 1.440 | L [o] 1.320 | LL L 22L 2L [o] 1.200 | L3 2LL L [o] 1.080 | 4L [o] 0.960 | L L2 [o] 0.840 | L L L L [o] 0.720 | L 2 2 [o] 0.600 | LL 2 2L [o] 0.480 | 3 LL [o] 0.360 | L LL LL [o] 0.240 | L [o] 0.120 | L L L [o] 0.000 | 22 L L [o] -0.120 | L L L [o] -0.240 | 22 L [o] -0.360 | L L L [o] -0.480 | LLL [o] -0.600 | L LLL L [o] -0.720 | [o] -0.840 | L [o] -0.960 | LL [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.480 0.640 0.800 0.960 1.120 1.280 1.440 [i] ? $cal e2=e*e $ [i] ? $yvar lnox $fit c+e+e2 $dis e $ [o] deviance = 3.4186 [o] d.f. = 84 [o] [o] estimate s.e. parameter [o] 1 -13.91 0.5143 1 [o] 2 0.02680 0.005116 C [o] 3 32.84 1.128 E [o] 4 -18.19 0.6170 E2 [o] scale parameter taken as 0.04070 [o] [i] ? $cal r=lnox-%fv $plot r %fv $ [o] 0.5500 | R [o] 0.5000 | [o] 0.4500 | R R [o] 0.4000 | R [o] 0.3500 | R R [o] 0.3000 | [o] 0.2500 | R [o] 0.2000 | R R R 2 [o] 0.1500 | R R RR 2R2 [o] 0.1000 | R R R R 323R [o] 0.0500 | R R R R R 2RR [o] 0.0000 | 2 RR R R R R [o] -0.0500 | R RR RR [o] -0.1000 | R R R R RRR R R [o] -0.1500 | R R R R [o] -0.2000 | R RR RR R [o] -0.2500 | R R 2 R R [o] -0.3000 | R R [o] -0.3500 | R R R [o] -0.4000 | R [o] -0.4500 | R [o] ----------:---------:---------:---------:---------:---------:---------: [o] -1.250 -0.750 -0.250 0.250 0.750 1.250 1.750 [i] ? $cal %s=%cu((nox-%exp(%fv))**2) $ [i] ? $print %s $ [o] 7.983 [i] ? $dis m $ [o] Current model: [o] [o] number of units is 88 [o] [o] y-variate LNOX [o] weight * [o] offset * [o] [o] probability distribution is NORMAL [o] link function is IDENTITY [o] scale parameter is to be estimated by the mean deviance [o] [o] terms = 1 + C + E + E2 [o] [i] ? $yvar nox $link log $fit . $dis e $ [w] -- model changed [o] deviance = 5.8391 at cycle 4 [o] d.f. = 84 [o] [o] estimate s.e. parameter [o] 1 -16.56 0.6553 1 [o] 2 0.02434 0.002869 C [o] 3 38.99 1.443 E [o] 4 -21.59 0.7907 E2 [o] scale parameter taken as 0.06951 [o] [i] ? $cal r=nox-%fv $plot r %fv $ [o] 0.8000 | [o] 0.7200 | R [o] 0.6400 | [o] 0.5600 | R [o] 0.4800 | R R [o] 0.4000 | R R [o] 0.3200 | RR R R RR RR [o] 0.2400 | 2 R R R R 2 R R [o] 0.1600 | RR R R R R [o] 0.0800 | RR 2 R R RR R R [o] 0.0000 | R2 R R R RR R R R2R R [o] -0.0800 | 2 R R R R R 2 [o] -0.1600 | R R R R R 2 [o] -0.2400 | R R RR R RR [o] -0.3200 | R R R [o] -0.4000 | R R R [o] -0.4800 | R R R [o] -0.5600 | [o] -0.6400 | R [o] -0.7200 | [o] -0.8000 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.000 0.800 1.600 2.400 3.200 4.000 4.800 [i] ? $plot r e $ [o] 0.8000 | [o] 0.7200 | R [o] 0.6400 | [o] 0.5600 | R [o] 0.4800 | R R [o] 0.4000 | R R [o] 0.3200 | R R R R R 2 R [o] 0.2400 | R R R R 3 RR R [o] 0.1600 | R R R R RR [o] 0.0800 | R R R R RR RRRR [o] 0.0000 | RR R RR RR 2R RR R R R [o] -0.0800 | R R RR RR R R R [o] -0.1600 | R R R R R R R [o] -0.2400 | R R R R R RR [o] -0.3200 | RRR [o] -0.4000 | R R R [o] -0.4800 | R RR [o] -0.5600 | [o] -0.6400 | R [o] -0.7200 | [o] -0.8000 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.480 0.640 0.800 0.960 1.120 1.280 1.440 [i] ? $cal e3=e**3 : e4=e**4 $fit +e3+e4 $dis e $ [o] deviance = 4.3856 (change = -1.453) at cycle 4 [o] d.f. = 82 (change = -2 ) [o] [o] estimate s.e. parameter [o] 1 43.61 10.39 1 [o] 2 0.02436 0.002528 C [o] 3 -236.5 48.06 E [o] 4 444.5 82.17 E2 [o] 5 -345.4 61.60 E3 [o] 6 94.68 17.09 E4 [o] scale parameter taken as 0.05348 [o] [i] ? $cal r=nox-%fv $plot r %fv $ [o] 0.8800 | [o] 0.8000 | [o] 0.7200 | R [o] 0.6400 | [o] 0.5600 | [o] 0.4800 | R R [o] 0.4000 | R R [o] 0.3200 | R R R RR R [o] 0.2400 | R R R R RR [o] 0.1600 | R 2 R R R R [o] 0.0800 | 22 RR R R R R R R [o] 0.0000 | 2RR R R R R RR RRR R [o] -0.0800 | R 2 RR RRR R [o] -0.1600 | R3 R RR RR R R RR R [o] -0.2400 | 2 R R R R R [o] -0.3200 | R R R R [o] -0.4000 | R R R [o] -0.4800 | [o] -0.5600 | R [o] -0.6400 | [o] -0.7200 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.000 0.800 1.600 2.400 3.200 4.000 4.800 [i] ? $plot r e $ [o] 0.8800 | [o] 0.8000 | [o] 0.7200 | R [o] 0.6400 | [o] 0.5600 | [o] 0.4800 | R R [o] 0.4000 | R R [o] 0.3200 | R R 2 RR [o] 0.2400 | R R R 3 [o] 0.1600 | R R R RR R R [o] 0.0800 | R R RR RR RR 2R R [o] 0.0000 | R R R RR R R R R R RR RR [o] -0.0800 | R R R R R R 2 R [o] -0.1600 | R RR R R R RR R R 2 R R [o] -0.2400 | R R RR R R R [o] -0.3200 | R R R R [o] -0.4000 | R RR [o] -0.4800 | [o] -0.5600 | R [o] -0.6400 | [o] -0.7200 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 0.480 0.640 0.800 0.960 1.120 1.280 1.440 [i] ? $plot r c $ [o] 0.8800 | [o] 0.8000 | [o] 0.7200 | R [o] 0.6400 | [o] 0.5600 | [o] 0.4800 | R R [o] 0.4000 | R R [o] 0.3200 | 3 R R R [o] 0.2400 | 2 R R R R [o] 0.1600 | 2 2 2 R [o] 0.0800 | 3 3 3 2 R [o] 0.0000 | R 3 5 3 2 [o] -0.0800 | 3 R 4 R [o] -0.1600 | 3 2 2 2 5 [o] -0.2400 | 3 R 2 R [o] -0.3200 | 2 R R [o] -0.4000 | 3 [o] -0.4800 | [o] -0.5600 | R [o] -0.6400 | [o] -0.7200 | [o] ----------:---------:---------:---------:---------:---------:---------: [o] 5.00 7.50 10.00 12.50 15.00 17.50 20.00 [i] ? $cal fc=c $factor fc 5 $fit -c+fc $dis e $ [f] ** factor level out of range, at [it -c+fc $] [f] [h] Unit 1 of the factor FC, which was declared as having 5 levels, does not, when [h] rounded, lie between 1 and 5. Edit the factor or reset its number of levels. [h] [i] ? $dis m $ [o] Current model: [o] [o] number of units is 88 [o] [o] y-variate NOX [o] weight * [o] offset * [o] [o] probability distribution is NORMAL [o] link function is LOGARITHM [o] scale parameter is to be estimated by the mean deviance [o] [o] terms = 1 + E + E2 + E3 + E4 + FC [o] [i] ? $fit -fc $dis e $ [o] deviance = 9.172 at cycle 4 [o] d.f. = 83 [o] [o] estimate s.e. parameter [o] 1 46.31 15.13 1 [o] 2 -244.9 69.84 E [o] 3 453.9 119.3 E2 [o] 4 -349.3 89.34 E3 [o] 5 95.01 24.77 E4 [o] scale parameter taken as 0.1105 [o] [i] ? $fit c+e+e2+e3+e4 $ [o] deviance = 4.3856 at cycle 4 [o] d.f. = 82 [o] [i] ? $cal fc=c/3-1 $factor fc 5 $ [i] ? $fit -c+fc $dis e $ [o] deviance = 4.2729 (change = -0.1127) at cycle 4 [o] d.f. = 79 (change = -3 ) [o] [o] estimate s.e. parameter [o] 1 43.37 10.46 1 [o] 2 -234.7 48.34 E [o] 3 441.1 82.63 E2 [o] 4 -342.7 61.93 E3 [o] 5 93.93 17.18 E4 [o] 6 0.1096 0.03417 FC(2) [o] 7 0.1698 0.03472 FC(3) [o] 8 0.2110 0.03395 FC(4) [o] 9 0.3103 0.03366 FC(5) [o] scale parameter taken as 0.05409 [o] [i] ? $fit c+e+e2+e3+e4 $ [o] deviance = 4.3856 at cycle 4 [o] d.f. = 82 [o] [i] ? $fit +fc $dis e $ [o] deviance = 4.2729 (change = -0.1127) at cycle 4 [o] d.f. = 79 (change = -3 ) [o] [o] estimate s.e. parameter [o] 1 43.22 10.46 1 [o] 2 0.02586 0.002805 C [o] 3 -234.7 48.34 E [o] 4 441.1 82.63 E2 [o] 5 -342.7 61.93 E3 [o] 6 93.93 17.18 E4 [o] 7 0.03205 0.03142 FC(2) [o] 8 0.01465 0.03119 FC(3) [o] 9 -0.02170 0.03180 FC(4) [o] 10 0.000 aliased FC(5) [o] scale parameter taken as 0.05409 [o] [i] ? $stop