35 |
|
* |
36 |
|
* [1] Meineke, et al., J. Comp. Chem. 26, 252-271 (2005). |
37 |
|
* [2] Fennell & Gezelter, J. Chem. Phys. 124, 234104 (2006). |
38 |
< |
* [3] Sun, Lin & Gezelter, J. Chem. Phys. 128, 24107 (2008). |
39 |
< |
* [4] Vardeman & Gezelter, in progress (2009). |
38 |
> |
* [3] Sun, Lin & Gezelter, J. Chem. Phys. 128, 234107 (2008). |
39 |
> |
* [4] Kuang & Gezelter, J. Chem. Phys. 133, 164101 (2010). |
40 |
> |
* [5] Vardeman, Stocker & Gezelter, J. Chem. Theory Comput. 7, 834 (2011). |
41 |
|
*/ |
42 |
|
|
43 |
|
#include "math/CubicSpline.hpp" |
43 |
– |
#include "utils/simError.h" |
44 |
|
#include <cmath> |
45 |
+ |
#include <cassert> |
46 |
+ |
#include <cstdio> |
47 |
|
#include <algorithm> |
46 |
– |
#include <iostream> |
48 |
|
|
49 |
|
using namespace OpenMD; |
50 |
|
using namespace std; |
51 |
|
|
52 |
< |
CubicSpline::CubicSpline() : generated(false), isUniform(true) {} |
52 |
> |
CubicSpline::CubicSpline() : generated(false), isUniform(true) { |
53 |
> |
data_.clear(); |
54 |
> |
} |
55 |
|
|
56 |
< |
void CubicSpline::addPoint(RealType xp, RealType yp) { |
57 |
< |
data.push_back(make_pair(xp, yp)); |
56 |
> |
void CubicSpline::addPoint(const RealType xp, const RealType yp) { |
57 |
> |
data_.push_back(make_pair(xp, yp)); |
58 |
|
} |
59 |
|
|
60 |
|
void CubicSpline::addPoints(const vector<RealType>& xps, |
61 |
|
const vector<RealType>& yps) { |
62 |
|
|
63 |
< |
if (xps.size() != yps.size()) { |
64 |
< |
printf( painCave.errMsg, |
65 |
< |
"CubicSpline::addPoints was passed vectors of different length!\n"); |
66 |
< |
painCave.severity = OPENMD_ERROR; |
64 |
< |
painCave.isFatal = 1; |
65 |
< |
simError(); |
66 |
< |
} |
67 |
< |
|
68 |
< |
for (int i = 0; i < xps.size(); i++) |
69 |
< |
data.push_back(make_pair(xps[i], yps[i])); |
63 |
> |
assert(xps.size() == yps.size()); |
64 |
> |
|
65 |
> |
for (unsigned int i = 0; i < xps.size(); i++) |
66 |
> |
data_.push_back(make_pair(xps[i], yps[i])); |
67 |
|
} |
68 |
|
|
69 |
|
void CubicSpline::generate() { |
70 |
|
// Calculate coefficients defining a smooth cubic interpolatory spline. |
71 |
|
// |
72 |
|
// class values constructed: |
73 |
< |
// n = number of data points. |
73 |
> |
// n = number of data_ points. |
74 |
|
// x = vector of independent variable values |
75 |
|
// y = vector of dependent variable values |
76 |
|
// b = vector of S'(x[i]) values. |
77 |
|
// c = vector of S"(x[i])/2 values. |
78 |
|
// d = vector of S'''(x[i]+)/6 values (i < n). |
79 |
|
// Local variables: |
80 |
< |
|
80 |
> |
|
81 |
|
RealType fp1, fpn, h, p; |
82 |
|
|
83 |
|
// make sure the sizes match |
84 |
|
|
85 |
< |
n = data.size(); |
85 |
> |
n = data_.size(); |
86 |
|
b.resize(n); |
87 |
|
c.resize(n); |
88 |
|
d.resize(n); |
92 |
|
bool sorted = true; |
93 |
|
|
94 |
|
for (int i = 1; i < n; i++) { |
95 |
< |
if ( (data[i].first - data[i-1].first ) <= 0.0 ) sorted = false; |
95 |
> |
if ( (data_[i].first - data_[i-1].first ) <= 0.0 ) sorted = false; |
96 |
|
} |
97 |
|
|
98 |
|
// sort if necessary |
99 |
|
|
100 |
< |
if (!sorted) sort(data.begin(), data.end()); |
100 |
> |
if (!sorted) sort(data_.begin(), data_.end()); |
101 |
|
|
102 |
|
// Calculate coefficients for the tridiagonal system: store |
103 |
|
// sub-diagonal in B, diagonal in D, difference quotient in C. |
104 |
|
|
105 |
< |
b[0] = data[1].first - data[0].first; |
106 |
< |
c[0] = (data[1].second - data[0].second) / b[0]; |
105 |
> |
b[0] = data_[1].first - data_[0].first; |
106 |
> |
c[0] = (data_[1].second - data_[0].second) / b[0]; |
107 |
|
|
108 |
|
if (n == 2) { |
109 |
|
|
110 |
|
// Assume the derivatives at both endpoints are zero. Another |
111 |
|
// assumption could be made to have a linear interpolant between |
112 |
|
// the two points. In that case, the b coefficients below would be |
113 |
< |
// (data[1].second - data[0].second) / (data[1].first - data[0].first) |
113 |
> |
// (data_[1].second - data_[0].second) / (data_[1].first - data_[0].first) |
114 |
|
// and the c and d coefficients would both be zero. |
115 |
|
b[0] = 0.0; |
116 |
< |
c[0] = -3.0 * pow((data[1].second - data[0].second) / |
117 |
< |
(data[1].first-data[0].first), 2); |
118 |
< |
d[0] = -2.0 * pow((data[1].second - data[0].second) / |
119 |
< |
(data[1].first-data[0].first), 3); |
116 |
> |
c[0] = -3.0 * pow((data_[1].second - data_[0].second) / |
117 |
> |
(data_[1].first-data_[0].first), 2); |
118 |
> |
d[0] = -2.0 * pow((data_[1].second - data_[0].second) / |
119 |
> |
(data_[1].first-data_[0].first), 3); |
120 |
|
b[1] = b[0]; |
121 |
|
c[1] = 0.0; |
122 |
|
d[1] = 0.0; |
123 |
< |
dx = 1.0 / (data[1].first - data[0].first); |
123 |
> |
dx = 1.0 / (data_[1].first - data_[0].first); |
124 |
|
isUniform = true; |
125 |
|
generated = true; |
126 |
|
return; |
129 |
|
d[0] = 2.0 * b[0]; |
130 |
|
|
131 |
|
for (int i = 1; i < n-1; i++) { |
132 |
< |
b[i] = data[i+1].first - data[i].first; |
132 |
> |
b[i] = data_[i+1].first - data_[i].first; |
133 |
|
if ( fabs( b[i] - b[0] ) / b[0] > 1.0e-5) isUniform = false; |
134 |
< |
c[i] = (data[i+1].second - data[i].second) / b[i]; |
134 |
> |
c[i] = (data_[i+1].second - data_[i].second) / b[i]; |
135 |
|
d[i] = 2.0 * (b[i] + b[i-1]); |
136 |
|
} |
137 |
|
|
138 |
|
d[n-1] = 2.0 * b[n-2]; |
139 |
|
|
140 |
|
// Calculate estimates for the end slopes using polynomials |
141 |
< |
// that interpolate the data nearest the end. |
141 |
> |
// that interpolate the data_ nearest the end. |
142 |
|
|
143 |
|
fp1 = c[0] - b[0]*(c[1] - c[0])/(b[0] + b[1]); |
144 |
|
if (n > 3) fp1 = fp1 + b[0]*((b[0] + b[1]) * (c[2] - c[1]) / |
145 |
|
(b[1] + b[2]) - |
146 |
< |
c[1] + c[0]) / (data[3].first - data[0].first); |
146 |
> |
c[1] + c[0]) / (data_[3].first - data_[0].first); |
147 |
|
|
148 |
|
fpn = c[n-2] + b[n-2]*(c[n-2] - c[n-3])/(b[n-3] + b[n-2]); |
149 |
< |
|
149 |
> |
|
150 |
|
if (n > 3) fpn = fpn + b[n-2] * |
151 |
< |
(c[n-2] - c[n-3] - (b[n-3] + b[n-2]) * |
152 |
< |
(c[n-3] - c[n-4])/(b[n-3] + b[n-4]))/(data[n-1].first - data[n-4].first); |
151 |
> |
(c[n-2] - c[n-3] - (b[n-3] + b[n-2]) * |
152 |
> |
(c[n-3] - c[n-4])/(b[n-3] + b[n-4])) / |
153 |
> |
(data_[n-1].first - data_[n-4].first); |
154 |
|
|
157 |
– |
|
155 |
|
// Calculate the right hand side and store it in C. |
156 |
|
|
157 |
|
c[n-1] = 3.0 * (fpn - c[n-2]); |
175 |
|
// Calculate the coefficients defining the spline. |
176 |
|
|
177 |
|
for (int i = 0; i < n-1; i++) { |
178 |
< |
h = data[i+1].first - data[i].first; |
178 |
> |
h = data_[i+1].first - data_[i].first; |
179 |
|
d[i] = (c[i+1] - c[i]) / (3.0 * h); |
180 |
< |
b[i] = (data[i+1].second - data[i].second)/h - h * (c[i] + h * d[i]); |
180 |
> |
b[i] = (data_[i+1].second - data_[i].second)/h - h * (c[i] + h * d[i]); |
181 |
|
} |
182 |
|
|
183 |
|
b[n-1] = b[n-2] + h * (2.0 * c[n-2] + h * 3.0 * d[n-2]); |
184 |
|
|
185 |
< |
if (isUniform) dx = 1.0 / (data[1].first - data[0].first); |
185 |
> |
if (isUniform) dx = 1.0 / (data_[1].first - data_[0].first); |
186 |
|
|
187 |
|
generated = true; |
188 |
|
return; |
197 |
|
// value of spline at t. |
198 |
|
|
199 |
|
if (!generated) generate(); |
203 |
– |
RealType dt; |
200 |
|
|
201 |
< |
if ( t < data[0].first || t > data[n-1].first ) { |
202 |
< |
sprintf( painCave.errMsg, |
207 |
< |
"CubicSpline::getValueAt was passed a value outside the range of the spline!\n"); |
208 |
< |
painCave.severity = OPENMD_ERROR; |
209 |
< |
painCave.isFatal = 1; |
210 |
< |
simError(); |
211 |
< |
} |
201 |
> |
assert(t > data_.front().first); |
202 |
> |
assert(t < data_.back().first); |
203 |
|
|
204 |
|
// Find the interval ( x[j], x[j+1] ) that contains or is nearest |
205 |
|
// to t. |
206 |
|
|
216 |
– |
int j; |
217 |
– |
|
207 |
|
if (isUniform) { |
208 |
|
|
209 |
< |
j = max(0, min(n-1, int((t - data[0].first) * dx))); |
209 |
> |
j = max(0, min(n-1, int((t - data_[0].first) * dx))); |
210 |
|
|
211 |
|
} else { |
212 |
|
|
213 |
|
j = n-1; |
214 |
|
|
215 |
|
for (int i = 0; i < n; i++) { |
216 |
< |
if ( t < data[i].first ) { |
216 |
> |
if ( t < data_[i].first ) { |
217 |
|
j = i-1; |
218 |
|
break; |
219 |
|
} |
222 |
|
|
223 |
|
// Evaluate the cubic polynomial. |
224 |
|
|
225 |
< |
dt = t - data[j].first; |
226 |
< |
return data[j].second + dt*(b[j] + dt*(c[j] + dt*d[j])); |
238 |
< |
|
225 |
> |
dt = t - data_[j].first; |
226 |
> |
return data_[j].second + dt*(b[j] + dt*(c[j] + dt*d[j])); |
227 |
|
} |
228 |
|
|
229 |
|
|
236 |
|
// pair containing value of spline at t and first derivative at t |
237 |
|
|
238 |
|
if (!generated) generate(); |
251 |
– |
RealType dt; |
239 |
|
|
240 |
< |
if ( t < data.front().first || t > data.back().first ) { |
241 |
< |
sprintf( painCave.errMsg, |
255 |
< |
"CubicSpline::getValueAndDerivativeAt was passed a value outside the range of the spline!\n"); |
256 |
< |
painCave.severity = OPENMD_ERROR; |
257 |
< |
painCave.isFatal = 1; |
258 |
< |
simError(); |
259 |
< |
} |
240 |
> |
assert(t > data_.front().first); |
241 |
> |
assert(t < data_.back().first); |
242 |
|
|
243 |
|
// Find the interval ( x[j], x[j+1] ) that contains or is nearest |
244 |
|
// to t. |
245 |
|
|
264 |
– |
int j; |
265 |
– |
|
246 |
|
if (isUniform) { |
247 |
|
|
248 |
< |
j = max(0, min(n-1, int((t - data[0].first) * dx))); |
248 |
> |
j = max(0, min(n-1, int((t - data_[0].first) * dx))); |
249 |
|
|
250 |
|
} else { |
251 |
|
|
252 |
|
j = n-1; |
253 |
|
|
254 |
|
for (int i = 0; i < n; i++) { |
255 |
< |
if ( t < data[i].first ) { |
255 |
> |
if ( t < data_[i].first ) { |
256 |
|
j = i-1; |
257 |
|
break; |
258 |
|
} |
261 |
|
|
262 |
|
// Evaluate the cubic polynomial. |
263 |
|
|
264 |
< |
dt = t - data[j].first; |
264 |
> |
dt = t - data_[j].first; |
265 |
|
|
266 |
< |
RealType yval = data[j].second + dt*(b[j] + dt*(c[j] + dt*d[j])); |
267 |
< |
RealType dydx = b[j] + dt*(2.0 * c[j] + 3.0 * dt * d[j]); |
266 |
> |
yval = data_[j].second + dt*(b[j] + dt*(c[j] + dt*d[j])); |
267 |
> |
dydx = b[j] + dt*(2.0 * c[j] + 3.0 * dt * d[j]); |
268 |
|
|
269 |
|
return make_pair(yval, dydx); |
270 |
|
} |