@Carl Love that is actually not a bad idea:D - i'll try tomorrow and tell how it went:)

@Preben Alsholm thanks anyway:) i might be able to use the residuals for something tho:)

@Preben Alsholm hello again, do you by chance know if u can get the LSSolve to give out the standard error of all the parameters individually or know if there is a general formula for how to calculate the Se(paramter1),Se(parameter2) and so on ?:)

@Preben Alsholm ye it seems like it lacks something before it can be considered a good model - either there is something wrong with the data obtained - or as you suggested i need to make new model. that the good thing about reseach you can always find more - thanks for helping out always good to have someelse look at the problem:)

@Preben Alsholm 

sorry:(

T := [0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58]

here is the correct values

@Preben Alsholm 

1. T := [0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98, 100] is the time data for S, P1 and P2_P2e.

2.yes i mean the inital point for the LSS solve. hmmm yes i could give a guess for the parameters i want to find - but i need to apply this in case where i only know that the parameters should be positive. if you leave out the inital point can the LSsolve stil make an accuate estimate of parameters or?

3. yes sorry forgot to wirte those, beta=10 and Tr=2

4. the explantion as to why k(t) is just multiplied is that the only effect it has is making up for the loss of signal (all data expect time is signal ammounts over time) - its sort of the noise of the data set (its acuttaly an NMR pulse). my professor told me that you can also divide the dataset with k(t) as well to reduce the noise (if that makes any sense?) - i study chemestry at DTU and this data is part of a project :)

thanks for the fast reply:)

@Preben Alsholm 

thanks for your help it works flawless now:D the data is not mesaured however when it is, its more or less just as accuarte :) its a metode from NMR messurering which gives a very excact signal value:)

do you know any good places where you can read up on maple coding ?:) - i think its really fun and i want to get better:)

@Preben Alsholm 

ups sorry posted wrong, the values for the other data's are:

Sd1:=[9999.99913146527, 8328.870587730016, 6937.009129218748, 5777.745632133724, 4812.209983843559, 4008.027767752663, 3338.234748901315, 2780.3727629912773, 2315.7366939901444, 1928.7472914685711, 1606.42879823157, 1337.9738730950157, 1114.3812206651082, 928.1537778450084, 773.0473372602229, 643.861178944546, 536.2636901656048, 446.64712642164017, 372.00664374480215, 309.8395462632648, 258.0613707922982, 214.9359947700637, 179.01742405677248, 149.10130874173097, 124.18456128296602, 103.43172297538936, 86.14695101495772, 71.75068688490619, 59.760223754895115, 49.77352131783417, 41.45572537575984, 34.527940176389414, 28.757877036724647, 23.952065701993305, 19.949367286742678, 16.61557128695749, 13.838895500983284, 11.526237971574666, 9.600055276660656, 7.995762497895868, 6.659567687926541, 5.546668226044262, 4.619748586020874, 3.847729146270299, 3.204724089933639, 2.669173453270794, 2.2231202199354003, 1.8516082220993855, 1.542180660048009, 1.2844624256040165, 1.0698122246826733]

P1d1:=[0., 67.86790056712294, 114.88787098501874, 145.95438088662502, 164.85650644237887, 174.5943243835592, 177.53238933657715, 175.52454334470625, 170.01528788038206, 162.12199855219257, 152.70149870604672, 142.40383375964961, 131.71564594259794, 120.99508533139499, 110.49982501393684, 100.4094855845229, 90.84352461449197, 81.875483096623, 73.54425405521964, 65.86297314169886, 58.825994886622986, 52.41433369396145, 46.59987753975884, 41.34861380141083, 36.62308863895286, 32.38424797333296, 28.59279488078898, 25.210168238042353, 22.19922762137446, 19.524715140537534, 17.15353915072512, 15.054929119094991, 13.200492993612231, 11.564203569809704, 10.122333808891973, 8.853355661845415, 7.737818276872426, 6.758211752777743, 5.898824382293833, 5.145598404644606, 4.485988231380927, 3.908824334321392, 3.4041829184889902, 2.963263879378151, 2.578277217202656, 2.242338207824242, 1.9493712043864035, 1.6940215802937413, 1.4715762013195168, 1.2778913607920699, 1.1093278786186604]

P2_P2ed1:=[0., 273.7455923151005, 468.9876833676771, 603.0127843540068, 689.7481005790843, 740.2955125551716, 763.4719502955126, 766.2500659106381, 754.1171464404207, 731.3670481166282, 701.3373064625542, 666.601455332508, 629.124726220671, 590.3898651600801, 551.4986101399664, 513.2533489361241, 476.2226694903023, 440.79380962636543, 407.2145259378454, 375.62638119929386, 346.0911052823648, 318.61137532326416, 293.14710951683435, 269.62817359442926, 247.96420137623375, 228.0521280575824, 209.7819044709044, 193.04077352346002, 177.71641462424827, 163.69919886815083, 150.88376071479496, 139.17003672588217, 128.4638987481884, 118.67748061285504, 109.72927725731319, 101.54407912344645, 94.05278617846177, 87.19214189455327, 80.9044148169022, 75.13704961088516, 69.8423036947357, 64.97688096589012, 60.50157299874978, 56.38091246020617, 52.58284331192939, 49.0784103341188, 45.8414694962003, 42.84842000063367, 40.07795717780249, 37.51084631990239, 35.129716406645024]

sorry for the inconvince 

hi agian, my dataset is larger than just the 5 points - the full set looks like this:

T1 := [0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98, 100]

Sd1 := [9999.99913146527, 8328.870587730016, 6937.009129218748, 5777.745632133724, 4812.209983843559, 4008.027767752663, 3338.234748901315, 2780.3727629912773, 2315.7366939901444, 1928.7472914685711, 1606.42879823157, 1337.9738730950157, 1114.3812206651082, 928.1537778450084, 773.0473372602229, 643.861178944546, 536.2636901656048, 446.64712642164017, 372.00664374480215, 309.8395462632648, 258.0613707922982, 214.9359947700637, 179.01742405677248, 149.10130874173097]

P1d1 := [0., 67.86790056712294, 114.88787098501874, 145.95438088662502, 164.85650644237887, 174.5943243835592, 177.53238933657715, 175.52454334470625, 170.01528788038206, 162.12199855219257, 152.70149870604672, 142.40383375964961, 131.71564594259794, 120.99508533139499, 110.49982501393684, 100.4094855845229, 90.84352461449197, 81.875483096623, 73.54425405521964, 65.86297314169886, 58.825994886622986, 52.41433369396145, 46.59987753975884, 41.34861380141083]

P2_P2ed1 := [0., 273.7455923151005, 468.9876833676771, 603.0127843540068, 689.7481005790843, 740.2955125551716, 763.4719502955126, 766.2500659106381, 754.1171464404207, 731.3670481166282, 701.3373064625542, 666.601455332508, 629.124726220671, 590.3898651600801, 551.4986101399664, 513.2533489361241, 476.2226694903023, 440.79380962636543, 407.2145259378454, 375.62638119929386, 346.0911052823648, 318.61137532326416, 293.14710951683435, 269.62817359442926]

the above code works flawless on the points spanning form T1[1] to T1[24]

the suddenly it says a signularity is present when ever i use LSSolve - which is wierd because i can make a numeric solution passed that suggested signularity with ease

the 24 elements gives the right parameters tho - i just find it odd that LSSolve cant evalute whenever you get above 

any chance someone knows why?:)

@Preben Alsholm 

hello yes i have an intial guess of the parameters which is: T1_p1=15, k1=0.02, k4=0.05 and keq=5

thanks for helping me out:D 

@Preben Alsholm 

thanks you've been a great help:D never heard aobut the LSSolve before, will def use it in future problems:D thanks agian for the fast reply - you've made my day if not my whole week:D

@Preben Alsholm 

thanks for the fast reply:)

i think i can identify the real from what you worte which is in turn - is there a way where you can actually make k and B into seperate paramenters? 

the model actually is the soultion to one of 2 differential eq's:

ode1 := diff(SS(t), t) = -k*SS(t)-SS(t)/B;
ode2 := diff(PP(t), t) = k*SS(t)-PP(t)/T1p;
 

where the inital values are PP(0)=0,SS(0)=100000

init:=PP(0)=0,SS(0)=100000

i then use dsolve({ode1,ode2,init})

to obtian:

ss := 1.00000*10^5*exp(-1.*t*k-1.*t/B)

and 

pp := 1.00000*10^5*T1p*k*B*exp(-1.*t/T1p)/(B*T1p*k-1.*B+T1p)-1.00000*10^5*B*T1p*k*exp(-(1.*(k*B+1.))*t/B)/((k*B+1.)*T1p-1.*B)

with the actual data : (sorry for type errors)

x:=<0,2,4,6,8>

y1:=<100000, 86089.76983, 74114.4849, 63804.98946, 54929.56828>

y2:=`<0, 8932.499583, 15965.05975, 21410.35044, 25533.98716>

is there a way i can find k, B and T1p form theses data as seperate parameters? (note they should all be postive)