how to fit exceeding probability curve? (Hazard curve)

Question

Mos_bad 2020-2-10

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/504765-how-to-fit-exceeding-probability-curve-hazard-curve

untitled.jpg

I want to fit a hazard curve (or exceedance= 1-CDF) to a set of raw data. Here is the code that I wrote.

As you may see in the attached file, I used linear polinomial fitting, but it does not seem accurate enogh.

yy=[216.623520000000	216.623520000000	234.746272000000	235.029440000000	263.063072000000	291.663040000000	325.643200000000	328.474880000000	334.138240000000	339.801600000000	342.633280000000	345.464960000000	365.286720000000	373.781760000000	385.108480000000	402.098560000000	410.593600000000	413.425280000000	424.752000000000	436.078720000000	438.910400000000	444.573760000000	447.405440000000	455.900480000000	461.563840000000	464.395520000000	470.058880000000	475.722240000000	489.880640000000	611.642880000000	620.137920000000	702.256640000000	832.513920000000	874.989120000000	880.652480000000	923.127680000000	999.583040000000	1064.71168000000	1098.69184000000	1127.00864000000	1152.49376000000	1166.65216000000	1192.13728000000	1282.75104000000	1446.98848000000	1452.65184000000	1458.31520000000	1489.46368000000	1509.28544000000	1523.44384000000	1529.10720000000	1557.42400000000	1560.25568000000	1571.58240000000	1585.74080000000	1588.57248000000	1591.40416000000	1599.89920000000	1749.97824000000	1846.25536000000	2069.95808000000	2137.91840000000	2177.56192000000	2231.36384000000	2234.19552000000	2242.69056000000	2304.98752000000]
POE_emp=[1	0.973684210526316	0.960526315789474	0.947368421052632	0.934210526315790	0.921052631578948	0.907894736842106	0.894736842105264	0.868421052631579	0.855263157894737	0.842105263157895	0.828947368421053	0.815789473684211	0.802631578947369	0.789473684210527	0.776315789473685	0.763157894736843	0.750000000000000	0.736842105263158	0.710526315789474	0.697368421052632	0.644736842105263	0.618421052631579	0.605263157894737	0.592105263157895	0.578947368421053	0.565789473684211	0.552631578947368	0.539473684210526	0.526315789473684	0.513157894736842	0.500000000000000	0.486842105263158	0.473684210526316	0.460526315789474	0.447368421052631	0.434210526315789	0.407894736842105	0.394736842105263	0.381578947368421	0.368421052631579	0.355263157894737	0.328947368421053	0.315789473684211	0.302631578947369	0.289473684210526	0.276315789473684	0.263157894736842	0.250000000000000	0.223684210526316	0.210526315789474	0.197368421052632	0.184210526315790	0.171052631578947	0.157894736842105	0.144736842105263	0.131578947368421	0.118421052631579	0.105263157894737	0.0921052631578947	0.0789473684210527	0.0657894736842105	0.0526315789473685	0.0394736842105263	0.0263157894736842	0.0131578947368421	0]
figure(1);
semilogx((POE_emp).*100,yy,'bs','MarkerFaceColor','blue','MarkerSize',8);hold on; 
c1=polyfit((POE_emp).*100,yy,1);
y_est=polyval(c1,(POE_emp).*100);
semilogx((POE_emp).*100,y_est,'-r');
set ( gca, 'xdir', 'reverse' );
% ylim([99 10000]); xlim([0.01 1]);
xlabel('Annual Probability of Exceedance','FontSize',14,'fontweight','bold','FontName','TMag_LHSe New Roman');
ylim([0 3000]); 
% xlim([0.01 0.98]);
xticks([1 10  100]); 
aa = get(gca,'XTickLabel');  
set(gca,'XTickLabel',aa,'fontsize',14,'FontWeight','bold');
grid minor; 