Getting error with best.fit()

[martik617]

below is the message after running the best.fit() function on my data:

Error in if (Di < EXPONENTIAL_EPS || Df < EXPONENTIAL_EPS || b < HARMONIC_EPS) Inf else if (abs(Df - :
missing value where TRUE/FALSE needed
In addition: Warning messages:
1: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1])), function(guess) sse(q, :
NA/NaN function evaluation
2: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1])), function(guess) sse(q, :
NA/NaN function evaluation
3: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1])), function(guess) sse(q, :
NA/NaN function evaluation
4: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1]), 1.5), function(guess) sse(q, :
NA/NaN function evaluation
5: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1]), 1.5), function(guess) sse(q, :
NA/NaN function evaluation
6: In nlminb(c(q[1], -((q[2] - q[1])/q[1])/(t[2] - t[1]), 1.5), function(guess) sse(q, :
NA/NaN function evaluation

[martik617]

Here is my code to reporduce the error:
q <- c(0,0,0,0,0,0,0,0,407.5,244.1,479.3,805.7,423.20000000000005,469.4,92.2,5.5,724,640.9,478.1,388.79999999999995,614.1,405.6,337.70000000000005,344.29999999999995,226.29999999999998,298.70000000000005,255.70000000000002,327.1
,271.5 ,232.6 ,248.9 ,189.39999999999998 ,207.7 ,211 ,165.6 ,115.30000000000001 ,126.4 ,106.2 ,0.1 ,84.3 ,39.5 ,0
,130.3 ,324.5 ,138.4 ,232.1 ,222.7 ,200.3 ,158.2 ,232.5 ,150.6 ,114.8 ,144.2 ,70.1 ,168 ,162 ,135.6
,119.2 ,136.1 ,135.10000000000002 ,0 ,212.4 ,161.39999999999998 ,34.400000000000006 ,179.5 ,114.8
,134.5 ,128.8 ,45.2 ,84.3 ,151.5 ,591.1 ,2194.4 ,2479.8 ,1896.6 ,1369.3 ,1300.6 ,4632.5999999999995 ,6836.7
,5690.6 ,7770.4000000000005 ,7261.4 ,5281.5999999999995 ,6062.1 ,5656.4 ,6497.5000000000009
,5746 ,12247 ,9802.1999999999989 ,10197.199999999997 ,10063.300000000003 ,8515.9 ,5992.0999999999995
,8891 ,6637.3 ,6427.3 ,1281.5 ,1073.1 ,9192.9999999999964 ,5779.7000000000007)

t <- seq(from = 30, to = 3000, by = 30)
df <- data.frame(t,q)
fit <- best.fit(q, t)

[derrickturk]

aRpsDCA doesn't currently do a good job of guessing initial declines from data like this, which begins with a series of zero rate values. This may be improved in future releases, especially because the initial guess here of (0 - 0) / 0 = NaN blows up the entire fitting process.

That said, this data needs to be filtered to get any meaningful DCA result. Dropping zero-rate observations will let best.fit run to completion, but looking at a plot of q vs. t shows that this data is not going to be meaningfully fit by a single Arps decline curve.