| 1 | #include <iostream> |
|---|
| 2 | #include <math.h> |
|---|
| 3 | int linear_regression(int num, float *x, float *y, int ilow, int ihigh, float &slope, float &errSlope, float &intercept, float &errIntercept, float &r) |
|---|
| 4 | { |
|---|
| 5 | /** |
|---|
| 6 | * Computes the linear best fit to data: $y=a x + b$, where $x$ and |
|---|
| 7 | * $y$ are arrays of size num, $a$ is the slope and $b$ the |
|---|
| 8 | * y-intercept. The values used in the arrays are those from ilow |
|---|
| 9 | * to ihigh. (ie. if the full arrays are being used, then ilow=0 |
|---|
| 10 | * and ihigh=num-1. Returns the values of slope & intercept (with |
|---|
| 11 | * errors) as well as r, the regression coefficient. |
|---|
| 12 | * \param num Size of the x & y arrays. |
|---|
| 13 | * \param x Array of abscissae. |
|---|
| 14 | * \param y Array of ordinates. |
|---|
| 15 | * \param ilow Minimum index of the arrays to be used (ilow=0 means |
|---|
| 16 | * start at the beginning). |
|---|
| 17 | * \param ihigh Maximum index of the arrays to be used (ihigh=num-1 |
|---|
| 18 | * means finish at the end). |
|---|
| 19 | * \param slope Returns value of the slope of the best fit line. |
|---|
| 20 | * \param errSlope Returns value of the estimated error in the slope |
|---|
| 21 | * value. |
|---|
| 22 | * \param intercept Returns value of the y-intercept of the best fit |
|---|
| 23 | * line. |
|---|
| 24 | * \param errIntercept Returns value of the estimated error in the |
|---|
| 25 | * value of the y-intercept. |
|---|
| 26 | * \param r Returns the value of the regression coefficient. |
|---|
| 27 | * \return If everything works, returns 0. If slope is infinite (eg, |
|---|
| 28 | * all points have same x value), returns 1. |
|---|
| 29 | */ |
|---|
| 30 | |
|---|
| 31 | if (ilow>ihigh) { |
|---|
| 32 | std::cerr << "Error! linear_regression.cc :: ilow (" << ilow |
|---|
| 33 | << ") > ihigh (" << ihigh << ")!!\n"; |
|---|
| 34 | return 1; |
|---|
| 35 | } |
|---|
| 36 | if (ihigh>num-1) { |
|---|
| 37 | std::cerr << "Error! linear_regression.cc :: ihigh (" <<ihigh |
|---|
| 38 | << ") out of bounds of array (>" << num-1 << ")!!\n"; |
|---|
| 39 | return 1; |
|---|
| 40 | } |
|---|
| 41 | if(ilow<0){ |
|---|
| 42 | std::cerr << "Error! linear_regression.cc :: ilow (" << ilow |
|---|
| 43 | << ") < 0. !!\n"; |
|---|
| 44 | return 1; |
|---|
| 45 | } |
|---|
| 46 | |
|---|
| 47 | double sumx,sumy,sumxx,sumxy,sumyy; |
|---|
| 48 | sumx=0.; |
|---|
| 49 | sumy=0.; |
|---|
| 50 | sumxx=0.; |
|---|
| 51 | sumxy=0.; |
|---|
| 52 | sumyy=0.; |
|---|
| 53 | int count=0; |
|---|
| 54 | for (int i=ilow;i<=ihigh;i++){ |
|---|
| 55 | count++; |
|---|
| 56 | sumx = sumx + x[i]; |
|---|
| 57 | sumy = sumy + y[i]; |
|---|
| 58 | sumxx = sumxx + x[i]*x[i]; |
|---|
| 59 | sumxy = sumxy + x[i]*y[i]; |
|---|
| 60 | sumyy = sumyy + y[i]*y[i]; |
|---|
| 61 | } |
|---|
| 62 | |
|---|
| 63 | const float SMALLTHING=1.e-6; |
|---|
| 64 | if(fabs(count*sumxx-sumx*sumx)<SMALLTHING) return 1; |
|---|
| 65 | else{ |
|---|
| 66 | |
|---|
| 67 | slope = (count*sumxy - sumx*sumy)/(count*sumxx - sumx*sumx); |
|---|
| 68 | errSlope = count / (count*sumxx - sumx*sumx); |
|---|
| 69 | |
|---|
| 70 | intercept = (sumy*sumxx - sumxy*sumx)/(count*sumxx - sumx*sumx); |
|---|
| 71 | errIntercept = sumxx / (count*sumxx - sumx*sumx); |
|---|
| 72 | |
|---|
| 73 | r = (count*sumxy - sumx*sumy) / |
|---|
| 74 | (sqrt(count*sumxx-sumx*sumx) * sqrt(count*sumyy-sumy*sumy) ); |
|---|
| 75 | |
|---|
| 76 | return 0; |
|---|
| 77 | |
|---|
| 78 | } |
|---|
| 79 | } |
|---|
