
Citation: Andres Fernandez, Norman R. Swanson. Further Evidence on the Usefulness of Real-Time Datasets for Economic Forecasting[J]. Quantitative Finance and Economics, 2017, 1(1): 2-25. doi: 10.3934/QFE.2017.1.2
[1] | Kai Wang, Guicang Zhang . Curve construction based on quartic Bernstein-like basis. AIMS Mathematics, 2020, 5(5): 5344-5363. doi: 10.3934/math.2020343 |
[2] | Jin Xie, Xiaoyan Liu, Lei Zhu, Yuqing Ma, Ke Zhang . The C3 parametric eighth-degree interpolation spline function. AIMS Mathematics, 2023, 8(6): 14623-14632. doi: 10.3934/math.2023748 |
[3] | Fengyi Chen . A new GM2(2,1) model based on a C1 convexity-preserving rational quadratic interpolation spline. AIMS Mathematics, 2024, 9(7): 17917-17931. doi: 10.3934/math.2024872 |
[4] | Shuangnian Hu, Yanyan Li, Rongquan Feng . Counting rational points of quartic diagonal hypersurfaces over finite fields. AIMS Mathematics, 2024, 9(1): 2167-2180. doi: 10.3934/math.2024108 |
[5] | Salwa Syazwani Mahzir, Md Yushalify Misro . Enhancing curve smoothness with whale optimization algorithm in positivity and monotonicity-preserving interpolation. AIMS Mathematics, 2025, 10(3): 6910-6933. doi: 10.3934/math.2025316 |
[6] | Lin Han, Guangyan Zhu, Zongbing Lin . On the rationality of generating functions of certain hypersurfaces over finite fields. AIMS Mathematics, 2023, 8(6): 13898-13906. doi: 10.3934/math.2023711 |
[7] | Sourav Shil, Hemant Kumar Nashine . Positive definite solution of non-linear matrix equations through fixed point technique. AIMS Mathematics, 2022, 7(4): 6259-6281. doi: 10.3934/math.2022348 |
[8] | Qi Xie, Yiting Huang . A class of generalized quadratic B-splines with local controlling functions. AIMS Mathematics, 2023, 8(10): 23472-23499. doi: 10.3934/math.20231193 |
[9] | Vediyappan Govindan, Inho Hwang, Choonkil Park . Hyers-Ulam stability of an n-variable quartic functional equation. AIMS Mathematics, 2021, 6(2): 1452-1469. doi: 10.3934/math.2021089 |
[10] | Salwa Syazwani Mahzir, Md Yushalify Misro, Kenjiro T. Miura . Preserving monotone or convex data using quintic trigonometric Bézier curves. AIMS Mathematics, 2024, 9(3): 5971-5994. doi: 10.3934/math.2024292 |
Shape-preserving interpolation is an important part in the application of engineering and sciences. The technique for finding the smoothness visual or graphical representation of data is not new; for example, it can be found in astronomy and meteorology. The main idea for the positivity preservation is to derive positive results in the entire given interval and if possible those rational splines should have free parameters. The free parameter provides the extra degree of freedom to the user in manipulating the final shape of the interpolating curves. Thus, shape-preserving interpolation research is focused on developing a new rational spline with the extra free parameters. Positivity preserving is about visualizing positive data sets, for instance, rainfall distribution at a certain location. The resulting interpolating curve must preserve the positivity of data sets everywhere and any negativity is meaningless.
Many researchers have contributed to the field of shape-preserving especially on the positivity preserving interpolation. A brief review is provided here. Asim and Brodlie [1] have preserved the shape-preserving for the positivity and constrained data by inserting one or more knots where are required to inserting. Goodman [2] have been surveyed the alternatives of the shape-preserving for interpolating curve. Sarfraz [3,4] have been used the rational function for preserve the shape-preserving interpolation.
Delbourgo & Gregory [5] give the clear concept of the shape-preserving interpolation by using a rational cubic spline. The amount of gas discharge during experiments by Brodlie and Butt [6], Brodlie et al. [7] and Sarfraz et al. [8] are the examples of positive data. Peng et al. [9] discussed the preserving of positivity interpolation of data on rectangular grids. Their scheme extended the rational cubic spline of Sarfraz [4] with two parameters to construct the bivariate rational interpolating function with C1 continuity. The main strategy of this scheme was by converting the bivariate interpolating function into bi-cubic. The sufficient condition for the positivity derived and used the lower bound have been discussed in Chan et al. [10].
Brodlie and Butt [6] and Butt and Brodlie [11] constructed the piecewise cubic Hermite interpolant with C1 continuity to preserve the positivity and monotonicity respectively. In their scheme, needs to insert one or two extra knots is needed to be in the interval where the shape violation exists. It will increase the total number of data points and increases the required computation. No derivation is required. Brodlie et al [7] extended the idea of Butt and Brodlie [11] to construct the positivity preserving of positive surface data arranged over a rectangular grid.
Karim and Pang [12] have proposed shape-preserving using rational cubic spline with three parameters (cubic/quadratic). Karim and Pang [13] have extended the function of rational cubic spline into C1 continuity interpolating curve for shape -preserving. Karim and Saaban [14] has been discussed shape-preserving interpolation with C1 continuity by using rational cubic ball function with one parameter and its application in image interpolation. This scheme has been extended the function of the univariate rational cubic ball to the bivariate function.
Hussain et al. [15] extended the application of the rational quartic spline given by Wang and Tan [16] to preserve the positivity of data. The data-dependent sufficient condition for the rational interpolant to satisfy the shape-preserving properties are derived. These schemes were not successful in producing positive and convex interpolating curves. The main difference between Wang and Tan [16] and Hussain et al. [15] scheme with the proposed scheme in this study is that we constructed the rational quartic spline with quadratic denominator with three parameters compared to their scheme which used two parameters to achieve the positive curve interpolation everywhere.
The main contribution of this study can be summarized as follow:
a. The proposed scheme does not require the modification of the first derivative if the shape-preserving is violated as shown in the work of Brodlie and Butt [6].
b. The proposed scheme is guaranteed to produce the positivity preserving interpolation on all intervals, while the schemes of Hussain et al. [15] and Sarfraz et al. [17] cannot not producing positive preserving interpolation on all intervals.
c. The proposed scheme is better than existing schemes, since it has extra free parameters to control the positive interpolating curve and many of the existing schemes are special cases of our proposed scheme and have similar results to existing schemes for various values of the shape parameters.
This paper is organized as follows: section 2 is devoted to the new rational quartic spline interpolation including construction and derivative estimation as in Harim et al. [18]. The sufficient condition for the positivity preserving is discussed in section 3. The numerical and graphical results for positivity preserving is discussed in section 4. The conclusion is given in the final section.
This section discusses the construction rational quartic spline with three parameters including the definition from Harim et al. [18].
Given scalar data set {(xI,fi),i=1,2,..n} where x1<x2<...<xn and the first derivative di, at the respective point. Then the rational quartic spline with three shape parameters αi>0,βi>0 and γi≥0 on the interval is given by ([18]):
S(x)=Pi(θ)Qi(θ) | (1) |
where
Pi(θ)=(1−θ)4αifi+(1−θ)3θAi+(1−θ)2θ2Bi+(1−θ)θ3Ci+θ4βifi+1Qi(θ)=αi(1−θ)2+γi(1−θ)θ+βiθ2 |
with hi=xi+1−xi, Δi=fi+1−fihi and θ=x−xihi.
The rational quartic spline in (1) is C1 continuity at the knots xi,i=1,2,...,n−1 and satisfy the following conditions:
S(xi)=fiS(xi+1)=fi+1S(1)(xi)=diS(1)(xi+1)=di+1 | (2) |
From condition in (2) and after some derivations, the unknown Ai,Bi and Ci are given as in Harim et al. [10].
Ai=(2αi+γi)fi+αihidiBi=(αi+γi)fi+1+(βi+γi)fiCi=(2βi+γi)fi+1−βihidi+1 | (3) |
where s(x)∈C(1)[x0,xn].
There are many alternative methods to estimate the values of first derivative such as Arithmetic Mean Method (AMM) by Gregory & Delbourgo [18], Geometric Mean Method (GMM) by Gregory & Delbourgo [19] and Harmonic Mean Method (HMM) by Gregory & Delbourgo [5]. For this study, we choose the Arithmetic Mean Method (AMM) because it is simple and suitable for all types of data sets. The derivative is given as follow Harim et al. [18].
Suppose a 2D set of data {(xi,fi),i=1,1,2,...,n} is considered that x1<x2<...<xn. The first derivatives, di at the point xi,i=1,2,...,n. Let hi=xi+1−xi and Δi=fi+1−fihi.
At the end points of x1 and xn.
d1=Δ1+(Δ1−Δ2)(h1h1+h2) | (4) |
dn=Δn−1+(Δn−1−Δn−2)(hn−1hn−1+hn−2) | (5) |
At the center point, xi=2,3,..,n−1 the value of di are given as
di=(hi−1Δi+hiΔi−1hi−1+hi) | (6) |
In this section, an effective and automatic method to generate the positive interpolating curve is described. The new rational quartic spline with three parameters in Eq (1) is used for this purpose. The data under consideration are positive. The sufficient condition is described in the following theorem.
In this study, given that given that {(xi,fi),i=1,1,2,...,n} with x1<x2<...<xn is positive to avoid dividing by zero such that
fi>0i=1,2,..,n | (7) |
The rational quartic spline, S(x)>0 if and only if for both polynomials Pi(θ)>0 and Qi(θ)>0. Since Qi(θ)>0 for αi>0,βi>0γi≥0i=1,2,…,n−1. Therefore, the sufficient condition for the positivity of S(x) can be derived from Equation (7). The main objective this paper is to find the best value of the parameter when αi>0,βi>0i=1,2,…,n−1 that will ensure that the rational quartic spline in Eq (1) is positive to interpolate of the positive interpolating curve on the whole domain in given data sets. We have adapted the method from Hussain et al. [15]. The following theorem gives sufficient conditions for the positivity preserving of the proposed scheme. It is data-dependent and has two free parameters to achieve the final interpolating curve of the positive result.
Theorem 1. For strictly positive data that has been discussed, the rational quartic spline interpolant defined on the interval [x1,xn] is positive in each subinterval [x1,xn],i=1,2,…n−1 the following sufficient conditions are satisfied:
γi>Max{0,[−2αifi−αihidifi],[−2βifi+1+βhidi+1fi+1]} | (8) |
Proof: The sufficient conditions of the positivity are derived from the condition in Eq (7) as follows:
By using Eq (7), it can be observed that αifi,βifi+1 and Bi are positive, so the Pi(θ) will be positive value if
Ai>0,Ci>0 | (9) |
Ai>0 If
γi>−2αifi−αihidifi | (10) |
Ci>0 If
γi>−2βifi+1+βihidi+1fi+1 | (11) |
The above constraints can be combined and rewritten as:
γi>Max{0,[−2αifi−αihidifi],[−2βifi+1+βhidi+1fi+1]} | (12) |
For the computer implementation, condition in Eq (12) can be rewritten as:
γi=λi+Max{0,[−2αifi−αihidifi],[−2βifi+1+βhidi+1fi+1]} | (13) |
where λi>0.
Algorithm for computer implementation as follow:
Step 1: Input the positive data for {(xi,fi),i=1,1,2,...,n}.
Step 2: Calculate the first derivative, di using Arithmetic Mean Method (AMM).
Step 3: Stimulate the value free parameters when αi>0,βi>0 and values of λi.
Step 4: Calculate the value of γi by using the sufficient condition in Equation (13).
Step 5: Construct the positive interpolating curve with C1 continuity.
Figure 1 shows flowchart for algorithm to obtain the positivity preserving interpolation curve.
This section is devoted to the numerical results for the positivity preserving interpolation including the comparison with the existing schemes. We test the scheme to four different positive data sets.
Table 1 shown positive data from demonstrates the level of oxygen from the experiment conducted. Data used Butt and Brodlie [11].
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 2 | 4 | 10 | 28 | 30 | 32 |
fi | 20.8 | 8.8 | 4.2 | 0.5 | 3.9 | 6.3 | 9.6 |
di | −7.85 | −4.15 | −1.8792 | −0.4153 | 1.0539 | 1.425 | 1.975 |
Figure 2 shows the positivity preserving interpolating curve for Table 1 using the existing scheme and the proposed scheme with a value of λi=0.1. Figure 2(a) shows the interpolating curve by the quartic polynomial when αi=1,βi=1,γi=2 and the curve gives the negative result as shown in Figure 3(a). Figure 2(b) and Figure 2(c) show the scheme of Hussain et al. [12] and Sarfraz et al. [17] and both schemes show the negative result on the interval [10,28] as shown in Figure 3(b) and Figure 3(c) respectively. Figure 2(d) until Figure 2(f) show the Karim and Pang [12], Qin et al. [21] and the proposed scheme using the same value of αi=1,βi=1 respectively. For Figure 2(e) not visual pleasing, because on interval [10,28] straight line compared with the others scheme. Figure 2(g) and Figure 2(h) show the interpolating the positivity curve by the proposed schemes with different values shown in Table 2. The values for γi obtained from Theorem 1.
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −7.85 | −4.15 | −1.879 | −0.415 | 1.054 | 1.425 | 1.975 | |
Figure 2(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0472 | 0.1 | 0.1 | - | |
Figure 2(g) | αi | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0074 | 0.1 | 0.1 | - | |
Figure 2(h) | αi | 1 | 1 | 1 | 1 | 2.5 | 1 | - |
βi | 3 | 6 | 2.5 | 9.5 | 2.5 | 1.5 | - | |
γi | 0.1 | 0.1 | 0.7845 | 32.4680 | 0.1 | 0.1 | - |
Figure 4 shows the interpolating curve for positivity preserving by the various value of λi=0.1,λi=0.5 (dashed) and λi=0.25 (dotted). From this figure, we can conclude that value of λi=0.5 more suitable to be chosen for this positive data.
Table 3 shows a positive data are taken from Hussain and Ali [22].
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
xi | 1 | 2 | 3 | 8 | 10 | 11 | 12 | 14 |
fi | 14 | 8 | 3 | 0.8 | 0.5 | 0.45 | 0.4 | 0.37 |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 |
Figure 5 shows the positivity preserving interpolation for positive data shown in Table 3. Figure 5(a) shows the interpolation curve for default quartic polynomial. Figure 5(b) and Figure 5(c) show curve for positivity used Hussain et al. [15] and Sarfraz et al. [17] schemes respectively. Figure 6 shows the zooming negative curve on the interval [3,8] for Figure 5(c) which is Sarfraz et al. [23] scheme. Figure 5(d) show the interpolating curve for positivity for Karim and Pang [12] meanwhile Figure 5(e) shows the positivity preserving interpolation for Qin et al. [21]. Figure 5(f) until Figure 5(h) show the interpolating curve for the proposed scheme by using values of different parameters shown in Table 4. The value of the parameter γi is calculated from Theorem 1 have been discussed in Eq (13). Figure 5(f) gives a loose curve compared to the other proposed schemes.
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 | |
Figure 5(f) | αi | 1 | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 5.17 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(g) | αi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - |
βi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - | |
γi | 0.1 | 0.1 | 1.37 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(h) | αi | 2.5 | 2.5 | 10 | 2.5 | 2.5 | 2.5 | 2.5 | - |
βi | 2.5 | 2.5 | 0.01 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 50.77 | 0.1 | 0.1 | 0.1 | 0.1 | - |
Figure 6 shows that Sarfraz et al. [17] fails to preserve the positivity of the positive data sets. Figure 7 shows the positivity interpolating curve by varying values of λi. Figure 7 shows the values of λi=1, λi=0.5 (dashed) and λi=2.5 (dotted). The figure above used the same value of parameters αi and βi but the different values of γi due to the calculation in Theorem 1. Curve with λi=2.5 gives a slightly different curve compared with λi=0.5 and λi=1.
Table 5 shows a positive data from Wu et al. [23].
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 0.25 | 0.5 | 1 | 1.2 | 1.8 | 2 |
fi | 2 | 0.8 | 0.5 | 0.1 | 1 | 0.5 | 1 |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 |
Figure 8 shows the shape-preserving for positivity interpolation curve for Table 5. Figure 8(a) shows the quartic polynomial interpolating curve with value αi=1,βi=1,γi=2. Figure 9 shows the zooming for Figure 8(a) that show the negative result for positivity interpolation. Figure 8(b) until Figure 8(e) show the positivity curve for Hussain et al. [15], Sarfraz et al. [17], Karim and Pang [12] and Qin et al. [21] respectively. For these figures, the curve nearly shows the negative result for the interval [3,7]. Figure 8(e) not visual pleasing curve interpolation on interval [1.2, 1.8] compared with the others scheme. Curve interpolation for the proposed scheme shown in Figure 8(f) until Figure 8(h) by using the values of different αi and βi shown in Table 6.
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 | |
Figure 8(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 13.01 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(g) | αi | 0.5 | 0.3 | 2.5 | 2.5 | 5.3 | 0.1 | - |
βi | 2.5 | 10.5 | 1.0 | 1.0 | 5.3 | 5.5 | - | |
γi | 0.1 | 0.1 | 13.03 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(h) | αi | 2.5 | 0.1 | 2.5 | 2.5 | 0.5 | 0.5 | - |
βi | 0.15 | 10.5 | 0.1 | 0.1 | 4.5 | 0.5 | - | |
γi | 0.1 | 0.1 | 1.393 | 0.1 | 0.1 | 0.1 | - |
Figure 10 shows the positivity interpolating curve by varying values of λi for all i=1,2,…n−1. Figure 10 shows the values of λi=0.1, λi=0.5 (dashed) and λi=2.5 (dotted). The Figure above used the same value of parameters αi and βi but different values of γi due to the calculation in Theorem 1. Interval [0.25, 0.5] with λi=2.5 gives smoother curve compared with others value. Curve with λi=0.1 is smoother or pleasing visually on the interval [1.2, 1.8]. For the other intervals, it gives the same curve for all three values of λi.
Table 7 shows a positive data from Hussain et al. [24].
i | 1 | 2 | 3 | 4 |
xi | 0 | 1 | 1.7 | 1.8 |
fi | 0.25 | 1 | 11.11 | 25 |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 |
Figure 11 shows the interpolating curve for positivity preserving interpolation curve by comparing the existing schemes and the proposed scheme of a rational quartic spline. Figure 11(a) is default quartic polynomial interpolating curve when αi=1,βi=1,γi=2 and that figure shows the negative result on the interval [0, 1] as shown in Figure 12(a). Figure 11(b) and Figure 11(c) show the scheme of Hussain et al. [15] and Sarfraz et al. [17], these schemes show the negative result on the interval [0, 1] as shown in Figure 12(b) and Figure 12(c) respectively. Figure 11(d) shows the scheme of Karim and Pang [12] meanwhile Figure 11(e) shows the interpolating curve from Qin et al. [21] scheme. The figure, not visual pleasing interpolating curve, because on interval [1.0, 1.7] have straight line compared with the others scheme. Figure 11(f) until Figure 11(h) show the curve interpolation for positivity preserving by the proposed schemes with the values of parameters as shown in Table 8. Figure 11(f) gives a more pleasing visually curve compared with other proposed schemes.
i | 1 | 2 | 3 | 4 | |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 | |
Figure 11(f) | αi | 1.5 | 1.5 | 1.5 | - |
βi | 0.1 | 0.1 | 1.0 | - | |
γi | 40.8772 | 0.5783 | 0.1 | - | |
Figure 11(g) | αi | 2.5 | 2.5 | 0.5 | - |
βi | 0.5 | 2.5 | 1.0 | - | |
γi | 68.062 | 14.5592 | 0.1 | - | |
Figure 11(h) | αi | 0.5 | 0.01 | 2.5 | - |
βi | 2.5 | 0.01 | 2.0 | - | |
γi | 17.0905 | 0.1578 | 0.1 | - |
Figure 13 shows the interpolating curve for positivity preserving by the various value of λi=0.1, λi=0.5 (dashed) and λi=2.5 (dotted). Value of λi=2.5 more converges to straight -line curve interpolation. For λi=0.1, give more pleasing curve interpolation.
Table 9 summarize of all presented schemes in this study for positivity preserving interpolation. From all four examples, Hussain et al. [15] fails to preserve the positivity for Examples 1 and 4. Meanwhile Sarfraz et al. [17] fails to preserve the positivity for Examples 1, 2 and 4. Finally, the proposed scheme preserves the positivity of the data for all examples.
Scheme | Example 1 | Example 2 | Example 3 | Example 4 |
Hussain et al. [15] | No | Yes | Yes | No |
Sarfraz et al. [17] | No | No | Yes | No |
The proposed scheme | Yes | Yes | Yes | Yes |
Final Remark: As seen in Figures 2 and 11, Qin et al. [21] shows interpolation curves that approaching to straight line. Even though the curves have no oscillation, however in terms of visually pleasing, their scheme is not. If we compare with other existing schemes, for instance Karim and Pang [12], the interpolating curves using [12] also show effect 'small oscillation' which not destroying the characteristic of the data. From Figures 4 and 13, we can see that, the proposed interpolating curves is approaching linear segments when λi=2.5 which is quite similar with Qin et al. [21] scheme. Furthermore, based on Figure 8(e), Qin et al. [21] is not visually pleasing since the interpolating curves are oscillated. Compared with our proposed schemes, all interpolating is visually pleasing. Without doubt, the proposed positivity-preserving using rational quartic spline provide good alternative to the existing schemes.
To achieve a smooth for positive data, a new C1 rational quartic spline (quartic/quadratic) with three parameters is proposed. The two free parameters are manipulated in Eq (13) to preserve the shape of the positivity preserving interpolation. This paper is an extension to the works of Hussain et al. [15] and Sarfraz et al. [17] that did not preserve the positivity results everywhere for positive data. From numerical and graphical results, we found that the proposed scheme preserves the positivity preserving interpolation everywhere than some established schemes such as Hussain et al. [15] and Sarfraz et al. [17]. By manipulating the value of λi,i=1,2,…,n−1, we found that the smallest value of λi, resulted in a smooth curve interpolation for positivity preserving. One possible extension to the present study is constructing positivity preserving interpolation by using the proposed rational quartic spline with C2 continuity. We may adopt the main idea in Zhu [25,26].
The authors would like to acknowledge Universiti Teknologi PETRONAS (UTP) and Ministry of Education, Malaysia for the financial support received in the form of a research grant: FRGS/ 1/2018/STG06/UTP/03/1/015MA0-020 and the author K.S. Nisar thanks to the Prince Sattam bin Abdulaziz University, Saudi Arabia for providing facilities and support.
The authors declare that there is no conflict of interests regarding the publication of this paper.
[1] | Amato J, Swanson NR (2001) The Real-Time Predictive Content of Money for Output. J Monet Econ 48: 324. |
[2] | Aruoba SB (2008) Real-Time Measurement of Business Conditions. J Bus Econ Stat, forthcom. |
[3] |
Aruoba SB, Diebold FX, Scotti C (2008) Data Revisions are not Well-Behaved. J Money, Credit and Banking 40: 319-340. doi: 10.1111/j.1538-4616.2008.00115.x
![]() |
[4] | Ashley R, Granger, Schmalensee R (1980) Advertising and Aggregate Consumption: An Analysis of Causality, Econometrica. 48: 1149-1167. |
[5] | Bernanke BS, Boivin J (2003) Monetary Policy in a Data-Rich Environment, J Monet Econ 50: 525546. |
[6] | Chao J, Corradi V, Swanson NR (2001) An out-of-sample Test for Granger Causality, Macroeconomic Dynamics 5: 598-620. |
[7] | Clark TE, McCrackenMW(2001) Tests of Equal Forecast Accuracy and Encompassing for Nested Models. J Econ 105: 85-110. |
[8] |
Clark TE, McCracken MW (2005) Evaluating Direct Multistep Forecasts. Econ Reviews 24: 369-404. doi: 10.1080/07474930500405683
![]() |
[9] | Clark TE, McCracken MW (2009) Tests of Equal Predictive Ability with Real-Time Data. J Business & Economic Stat 27: 441-454. |
[10] |
Corradi V, Swanson NR (2002) A Consistent Test for Out of Sample Nonlinear Predictive Ability. J Econ 110:353-381. doi: 10.1016/S0304-4076(02)00099-4
![]() |
[11] | Corradi VA, Fernandez, Swanson NR (2009) Information in the Revision Process of Real-Time Datasets. J Business & Economic Stat 27: 455-467. |
[12] | Corradi V, Swanson NR (2006) Predictive Density Evaluation, in: Handbook of Economic Forecasting, eds. Clive W.J. Granger, Graham Elliot and Allan Timmermann, Elsevier, Amsterdam, 197-284. |
[13] | Croushore D (2006) Forecasting with Real-Time Macroeconomic Data, in: Handbook of Economic Forecasting, eds. CliveW.J. Granger, Graham Elliot and Allan Timmermann, Elsevier, Amsterdam, 961982. |
[14] | Croushore D, Stark T (2001) A Real-Time Dataset for Macroeconomists. J Econ 105: 111130. |
[15] | Croushore D, Stark T (2003) A Real-Time Dataset for Macroeconomists: Does Data Vintage Matter? Rev Econ and Stat 85: 605617. |
[16] | Diebold FX, Mariano RS (1995) Comparing Predictive Accuracy. J Business and Economic Stat 13: 253-263. |
[17] | Diebold FX, Rudebusch GD (1991) Forecasting Output with the Composite Leading Index: A Real-Time Analysis. J American Stat Assoc 86: 603610. |
[18] | Faust J, Wright J (2009) Comparing Greenbook and Reduced Form Forecasts using a Large Realtime Dataset. J Business & Economic Stat. |
[19] |
Franses PH (2013) Data Revisions and Periodic Properties of Macroeconomic Data. Econ Letters 120: 139-141. doi: 10.1016/j.econlet.2013.04.014
![]() |
[20] | Franses PH, Segers R (2010) Seasonality in Revisions in Macroeconomic Data. J O cial Stat 26: 361-369. |
[21] | Gallo GM, Marcellino M (1999) Ex Post and Ex Ante Analysis of Provisional Data. J Forec 18: 421433. |
[22] |
Garratt A, Koop G, Mise E, et al. (2009) Real-Time Prediction with UK Monetary Aggregates in the Presence of Model Uncertainty. J Business and Economic Stat 27: 480-491. doi: 10.1198/jbes.2009.07208
![]() |
[23] | Ghysels E, Swanson NR, Callan M (2002), Monetary Policy Rules with Model and Data Uncertainty. Southern Econ J 69, 239-265. |
[24] | Gilbert T (2011), Information Aggregation Around Macroeconomic Announcements: Revisions Matter. J Financial Econ 101: 114-131. |
[25] | Hamilton JD, Perez-Quiros G (1996) What Do the Leading Indicators Lead? J Business 69: 2749. |
[26] | Kavajecz KA, Collins S (1995) Rationality of Preliminary Money Stock Estimates. Rev Econ and Stat 77: 3241. |
[27] | Keane MP, Runkle DE (1990) Testing the Rationality of Price Forecasts: New Evidence from Panel Data. American Econ Rev 80: 714735. |
[28] | Mankiw NG, Runkle DE, Shapiro MD(1984) Are Preliminary Announcements of the Money Stock Rational Forecasts? J Monetary Econ 14: 1527. |
[29] | Mankiw NG, Shapiro MD (1986) News or Noise: an Analysis of GNP Revisions. Surv of Current Bus 66: 2025. |
[30] | Mariano RS, Tanizaki H (1995) Prediction of Final Data with Use of Preliminary and/or Revised Data. J Forec 14: 351380. |
[31] | Mork KA (1987) Aint Behavin: Forecast Errors and Measurement Errors in Early GNP Estimates. J Business & Economic Stat 5: 165175. |
[32] |
Rathjens P, Robins RP (1995) Do Government Agencies Use Public Data: the Case of GNP. Rev Econ and Stat 77: 170-172. doi: 10.2307/2110002
![]() |
[33] | Robertson JC, Tallman EW (1998) Data Vintages and Measuring Forecast Performance, Federal Reserve Bank of Atlanta Economic Review 83 (Fourth Quarter), 420. |
[34] |
Stock JH, Watson MW (1989) Interpreting the Evidence on Money-Income Causality. J Econ 40: 161-181. doi: 10.1016/0304-4076(89)90035-3
![]() |
[35] | Swanson NR, Ghysels E, Callan M (1999) A Multivariate Time Series Analysis of the Data Revision Process for Industrial Production and the Composite Leading Indicator, in R.F. Engle and H. White, eds., Cointegration, Causality, and Forecasting: A Festschrift in Honour of Clive W.J. Granger, Oxford: Oxford University Press, 45-75. |
[36] | Swanson NR, Dijk D(2006) Are Statistical Reporting Agencies Getting It Right? Data Rationality and Business Cycle Asymmetry. J Business & Economic Stat 24: 2442. |
1. | Samsul Ariffin Abdul Karim, Rational Bi-Quartic Spline With Six Parameters for Surface Interpolation With Application in Image Enlargement, 2020, 8, 2169-3536, 115621, 10.1109/ACCESS.2020.3002387 | |
2. | Kottakkaran Sooppy Nisar, Vinita Sharma, Asif Khan, Lupaş blending functions with shifted knots and q-Bézier curves, 2020, 2020, 1029-242X, 10.1186/s13660-020-02450-5 | |
3. | Chengzhi Liu, Zhongyun Liu, Xuli Han, Preconditioned progressive iterative approximation for tensor product Bézier patches, 2021, 185, 03784754, 372, 10.1016/j.matcom.2021.01.002 | |
4. | Sidra Maqsood, Muhammad Abbas, Kenjiro T. Miura, Abdul Majeed, Azhar Iqbal, Geometric modeling and applications of generalized blended trigonometric Bézier curves with shape parameters, 2020, 2020, 1687-1847, 10.1186/s13662-020-03001-4 | |
5. | Saumya Ranjan Jena, Guesh Simretab Gebremedhin, Octic B-spline Collocation Scheme for Numerical Investigation of Fifth Order Boundary Value Problems, 2022, 8, 2349-5103, 10.1007/s40819-022-01437-8 | |
6. | Samsul Ariffin Abdul Karim, Azizan Saaban, Van Thien Nguyen, Construction of C1 Rational Bi-Quartic Spline With Positivity-Preserving Interpolation: Numerical Results and Analysis, 2021, 9, 2296-424X, 10.3389/fphy.2021.555517 | |
7. | Hrushikesh Jena, Mahendra Kumar Jena, Construction of Trigonometric Box Splines and the Associated Non-Stationary Subdivision Schemes, 2021, 7, 2349-5103, 10.1007/s40819-021-01069-4 | |
8. | Dominik Prossel, Uwe D. Hanebeck, 2024, Spline-Based Density Estimation Minimizing Fisher Information, 978-1-7377497-6-9, 1, 10.23919/FUSION59988.2024.10706290 | |
9. | Salwa Syazwani Mahzir, Md Yushalify Misro, Shape preserving interpolation of positive and range-restricted data using quintic trigonometric Bézier curves, 2023, 80, 11100168, 122, 10.1016/j.aej.2023.08.009 | |
10. | Ayesha Munir, Shamaila Samreen, Abdul Ghaffar, Mustafa Inc, Muhammad Sarfraz, A new cubic trigonometric Nu-Spline with shape control parameter and its applications, 2024, 139, 2190-5444, 10.1140/epjp/s13360-024-05381-y | |
11. | Salwa Syazwani Mahzir, Md Yushalify Misro, 2024, 3203, 0094-243X, 030003, 10.1063/5.0224390 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 2 | 4 | 10 | 28 | 30 | 32 |
fi | 20.8 | 8.8 | 4.2 | 0.5 | 3.9 | 6.3 | 9.6 |
di | −7.85 | −4.15 | −1.8792 | −0.4153 | 1.0539 | 1.425 | 1.975 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −7.85 | −4.15 | −1.879 | −0.415 | 1.054 | 1.425 | 1.975 | |
Figure 2(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0472 | 0.1 | 0.1 | - | |
Figure 2(g) | αi | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0074 | 0.1 | 0.1 | - | |
Figure 2(h) | αi | 1 | 1 | 1 | 1 | 2.5 | 1 | - |
βi | 3 | 6 | 2.5 | 9.5 | 2.5 | 1.5 | - | |
γi | 0.1 | 0.1 | 0.7845 | 32.4680 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
xi | 1 | 2 | 3 | 8 | 10 | 11 | 12 | 14 |
fi | 14 | 8 | 3 | 0.8 | 0.5 | 0.45 | 0.4 | 0.37 |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 | |
Figure 5(f) | αi | 1 | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 5.17 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(g) | αi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - |
βi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - | |
γi | 0.1 | 0.1 | 1.37 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(h) | αi | 2.5 | 2.5 | 10 | 2.5 | 2.5 | 2.5 | 2.5 | - |
βi | 2.5 | 2.5 | 0.01 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 50.77 | 0.1 | 0.1 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 0.25 | 0.5 | 1 | 1.2 | 1.8 | 2 |
fi | 2 | 0.8 | 0.5 | 0.1 | 1 | 0.5 | 1 |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 | |
Figure 8(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 13.01 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(g) | αi | 0.5 | 0.3 | 2.5 | 2.5 | 5.3 | 0.1 | - |
βi | 2.5 | 10.5 | 1.0 | 1.0 | 5.3 | 5.5 | - | |
γi | 0.1 | 0.1 | 13.03 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(h) | αi | 2.5 | 0.1 | 2.5 | 2.5 | 0.5 | 0.5 | - |
βi | 0.15 | 10.5 | 0.1 | 0.1 | 4.5 | 0.5 | - | |
γi | 0.1 | 0.1 | 1.393 | 0.1 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 |
xi | 0 | 1 | 1.7 | 1.8 |
fi | 0.25 | 1 | 11.11 | 25 |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 |
i | 1 | 2 | 3 | 4 | |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 | |
Figure 11(f) | αi | 1.5 | 1.5 | 1.5 | - |
βi | 0.1 | 0.1 | 1.0 | - | |
γi | 40.8772 | 0.5783 | 0.1 | - | |
Figure 11(g) | αi | 2.5 | 2.5 | 0.5 | - |
βi | 0.5 | 2.5 | 1.0 | - | |
γi | 68.062 | 14.5592 | 0.1 | - | |
Figure 11(h) | αi | 0.5 | 0.01 | 2.5 | - |
βi | 2.5 | 0.01 | 2.0 | - | |
γi | 17.0905 | 0.1578 | 0.1 | - |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 2 | 4 | 10 | 28 | 30 | 32 |
fi | 20.8 | 8.8 | 4.2 | 0.5 | 3.9 | 6.3 | 9.6 |
di | −7.85 | −4.15 | −1.8792 | −0.4153 | 1.0539 | 1.425 | 1.975 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −7.85 | −4.15 | −1.879 | −0.415 | 1.054 | 1.425 | 1.975 | |
Figure 2(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0472 | 0.1 | 0.1 | - | |
Figure 2(g) | αi | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
γi | 0.1 | 0.1 | 0.7845 | 13.0074 | 0.1 | 0.1 | - | |
Figure 2(h) | αi | 1 | 1 | 1 | 1 | 2.5 | 1 | - |
βi | 3 | 6 | 2.5 | 9.5 | 2.5 | 1.5 | - | |
γi | 0.1 | 0.1 | 0.7845 | 32.4680 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
xi | 1 | 2 | 3 | 8 | 10 | 11 | 12 | 14 |
fi | 14 | 8 | 3 | 0.8 | 0.5 | 0.45 | 0.4 | 0.37 |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
di | −6.5 | −5.5 | −4.24 | −0.232 | −0.083 | −0.05 | −0.038 | 0.0083 | |
Figure 5(f) | αi | 1 | 1 | 1 | 1 | 1 | 1 | 1 | - |
βi | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 5.17 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(g) | αi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - |
βi | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | 0.25 | - | |
γi | 0.1 | 0.1 | 1.37 | 0.1 | 0.1 | 0.1 | 0.1 | - | |
Figure 5(h) | αi | 2.5 | 2.5 | 10 | 2.5 | 2.5 | 2.5 | 2.5 | - |
βi | 2.5 | 2.5 | 0.01 | 2.5 | 2.5 | 2.5 | 2.5 | - | |
γi | 0.1 | 0.1 | 50.77 | 0.1 | 0.1 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
xi | 0 | 0.25 | 0.5 | 1 | 1.2 | 1.8 | 2 |
fi | 2 | 0.8 | 0.5 | 0.1 | 1 | 0.5 | 1 |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 |
i | 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
di | −6.6 | −3 | −1.067 | 2.986 | 3.167 | 1.667 | 3.333 | |
Figure 8(f) | αi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - |
βi | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | - | |
γi | 0.1 | 0.1 | 13.01 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(g) | αi | 0.5 | 0.3 | 2.5 | 2.5 | 5.3 | 0.1 | - |
βi | 2.5 | 10.5 | 1.0 | 1.0 | 5.3 | 5.5 | - | |
γi | 0.1 | 0.1 | 13.03 | 0.1 | 0.1 | 0.1 | - | |
Figure 8(h) | αi | 2.5 | 0.1 | 2.5 | 2.5 | 0.5 | 0.5 | - |
βi | 0.15 | 10.5 | 0.1 | 0.1 | 4.5 | 0.5 | - | |
γi | 0.1 | 0.1 | 1.393 | 0.1 | 0.1 | 0.1 | - |
i | 1 | 2 | 3 | 4 |
xi | 0 | 1 | 1.7 | 1.8 |
fi | 0.25 | 1 | 11.11 | 25 |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 |
i | 1 | 2 | 3 | 4 | |
di | −7.2962 | 8.7962 | 123.4286 | 154.5714 | |
Figure 11(f) | αi | 1.5 | 1.5 | 1.5 | - |
βi | 0.1 | 0.1 | 1.0 | - | |
γi | 40.8772 | 0.5783 | 0.1 | - | |
Figure 11(g) | αi | 2.5 | 2.5 | 0.5 | - |
βi | 0.5 | 2.5 | 1.0 | - | |
γi | 68.062 | 14.5592 | 0.1 | - | |
Figure 11(h) | αi | 0.5 | 0.01 | 2.5 | - |
βi | 2.5 | 0.01 | 2.0 | - | |
γi | 17.0905 | 0.1578 | 0.1 | - |
Scheme | Example 1 | Example 2 | Example 3 | Example 4 |
Hussain et al. [15] | No | Yes | Yes | No |
Sarfraz et al. [17] | No | No | Yes | No |
The proposed scheme | Yes | Yes | Yes | Yes |