|
|
|
|
| Function | new_linreg_struct |
| Include file | LINREG_R.H |
| Prototype | LINREG *new_linreg_struct( void ) |
| Remarks | Allocate and return a virgin linear regression structure. |
| Function | calc_linreg |
| Include file | LINREG_R.H |
| Prototype | void calc_linreg( LINREG *linreg , int data_max , float *data_ptr ) |
| Remarks | Calculate a least-squares linear regression on the supplied data. The parameter data_ptr is a pointer to a data series that is data_max items long. Least-squares regression uses a sequential integer count as the x-axis (or as the independent variable) to generate the projection of the dependent variable (the inputed price series). So this means least squares uses time (or bar periods if you prefer) to project prices. |
| Function | calc_best_fit_linreg |
| Include file | LINREG_R.H |
| Prototype | void calc_best_fit_linreg( LINREG *linreg , int data_max , float *data_ptr_ind , float *data_ptr_dep ) |
| Remarks | Calculate a best-fit linear regression on the supplied data. The parameters data_ptr_ind & data_ptr_dep are pointers to data series that are data_max items long. A best-fit linear regression uses one data series as the x-axis (or as the independent variable) to generate the projection of the dependent variable. This means best-fit can be used to project the prices of one series given the past relationship of 2 prices series. For example given the past prices of corn & soybeans - calc the regression, then given a new price for corn you could project a new price for beans. |
| Function | finish_regression_calcs |
| Include file | LINREG_R.H |
| Prototype | void finish_regression_calcs( LINREG *linreg , int data_max , float *data_ptr , double sum_x , double sum_y , double sum_xy , double sum_x_sqed , double sum_y_sqed ) |
| Remarks | This routine actually does the regression calcs after the input data has been processed (for both regression styles). |
| Function | calc_new_price_with_linreg |
| Include file | LINREG_R.H |
| Prototype | float calc_new_price_with_linreg( LINREG *linreg , float cur_dep_data ) |
| Remarks | Once you have calculated a regression on a data series, use this routine to project the new data from the regression. This routine can use the regression output from either least-squares or best-fit regression. Just pass a bar number as the current dependent data value or pass an actual price for a best_fit projection. |
| Function | detrend_data_series |
| Include file | LINREG_R.H |
| Prototype | void detrend_data_series( int data_cnt , float *data_ptr ) |
| Remarks | Use a least-squares regression to remove the trend from a data series. |
| Function | restore_trend_2_data_series |
| Include file | LINREG_R.H |
| Prototype | void restore_trend_2_data_series( LINREG *linreg , int data_cnt , float *data_ptr ) |
| Remarks | Use a regression calc to restore the trend to a data series. |
| Function | new_tslr_struct |
| Include file | LINREG_R.H |
| Prototype | TSLR *new_tslr_struct( int tslr_length ) |
| Remarks | Allocate, initialize and return a virgin Time Series Linear Regression indicator structure. |
| Function | reset_tslr_struct |
| Include file | LINREG_R.H |
| Prototype | void reset_tslr_struct( TSLR *tslr_ptr ) |
| Remarks | Reset a TSLR data structure for a new data series. |
| Function | update_tslr_struct |
| Include file | LINREG_R.H |
| Prototype | void update_tslr_struct( TSLR *tslr_ptr , float data ) |
| Remarks | Update the caller's TSLR with the new data. Once we have collected enough data to calc the regression, tslr_ptr->enough_data will be set to TRUE and the tslr var can be used. |
| Function | free_tslr_struct |
| Include file | LINREG_R.H |
| Prototype | void free_tslr_struct( TSLR *tslr_ptr ) |
| Remarks | Return the TSLR and its sub-structure memory to the system. |
| Function | new_tslr_ind_struct |
| Include file | LINREG_R.H |
| Prototype | TSLR_IND *new_tslr_ind_struct( int data_cnt ) |
| Remarks | Allocate, and return a virgin TSLR indicator display structure for data_cnt items. |
| Function | free_tslr_ind_struct |
| Include file | LINREG_R.H |
| Prototype | void free_tslr_ind_struct( TSLR_IND *tslr_ptr ) |
| Remarks | Return the TSLR indicator display structure and its sub-structure memory to the system. |
| Function | calc_tslr_ind |
| Include file | LINREG_R.H |
| Prototype | void calc_tslr_ind( DATA_REC *data_ptr , int data_cnt , int which_field , TSLR_IND *inds , int tslr_length ) |
| Remarks | Calculate a Time Series least-squares Linear Regression indicator for the requested field of the DATA_REC data series. The indicators will be built in the caller's TSLR_IND indicator display structure inds. |
