Per #2583, remove the DS_ADD_OERR and DS_MULT_OERR ECNT columns and r…

…ename DS_OERR as DSS, since observation error is not actually involved in its computation.

data/table_files/met_header_columns_V12.0.txt

@@ -19,7 +19,7 @@ V12.0 : STAT : PJC      : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID
 V12.0 : STAT : PRC      : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL (N_THRESH) THRESH_[0-9]* PODY_[0-9]* POFD_[0-9]*
 V12.0 : STAT : PSTD     : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL (N_THRESH) BASER BASER_NCL BASER_NCU RELIABILITY RESOLUTION UNCERTAINTY ROC_AUC BRIER BRIER_NCL BRIER_NCU BRIERCL BRIERCL_NCL BRIERCL_NCU BSS BSS_SMPL THRESH_[0-9]*
 V12.0 : STAT : ECLV     : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL BASER VALUE_BASER (N_PTS) CL_[0-9]* VALUE_[0-9]*
-V12.0 : STAT : ECNT     : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_ENS CRPS CRPSS IGN ME RMSE SPREAD ME_OERR RMSE_OERR SPREAD_OERR SPREAD_PLUS_OERR CRPSCL CRPS_EMP CRPSCL_EMP CRPSS_EMP CRPS_EMP_FAIR SPREAD_MD MAE MAE_OERR BIAS_RATIO N_GE_OBS ME_GE_OBS N_LT_OBS ME_LT_OBS IGN_OERR_CONV IGN_OERR_CORR DS_OERR DS_ADD_OERR DS_MULT_OERR
+V12.0 : STAT : ECNT     : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_ENS CRPS CRPSS IGN ME RMSE SPREAD ME_OERR RMSE_OERR SPREAD_OERR SPREAD_PLUS_OERR CRPSCL CRPS_EMP CRPSCL_EMP CRPSS_EMP CRPS_EMP_FAIR SPREAD_MD MAE MAE_OERR BIAS_RATIO N_GE_OBS ME_GE_OBS N_LT_OBS ME_LT_OBS IGN_OERR_CONV IGN_OERR_CORR DSS
 V12.0 : STAT : RPS      : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_PROB RPS_REL RPS_RES RPS_UNC RPS RPSS RPSS_SMPL RPS_COMP
 V12.0 : STAT : RHIST    : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL (N_RANK) RANK_[0-9]*
 V12.0 : STAT : PHIST    : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL BIN_SIZE (N_BIN) BIN_[0-9]*

docs/Users_Guide/appendixC.rst

@@ -1094,12 +1094,12 @@ Called "IGN_CONV_OERR" and "IGN_CORR_OERR" in ECNT output :numref:`table_ES_head
 
 TODO: Eric add 2 equations and interpretation advice here
 
-Dawid-Sebastiani Scoring Rules
-------------------------------
+Dawid-Sebastiani Scoring Rule
+-----------------------------
 
-Called "DS_OERR", "DS_ADD_OERR", and "DS_MULT_OERR" in ECNT output :numref:`table_ES_header_info_es_out_ECNT`
+Called "DSS" in ECNT output :numref:`table_ES_header_info_es_out_ECNT`
 
-TODO: Eric add 3 equations and interpretation advice here
+TODO: Eric add equation 16 and interpretation advice here
 
 IGN
 ---

docs/Users_Guide/ensemble-stat.rst

@@ -656,14 +656,8 @@ The format of the STAT and ASCII output of the Ensemble-Stat tool are described
     - IGN_CORR_OERR
     - Error-corrected logarithmic scoring rule (i.e. ignornance score) from Equation 7 of :ref:`Ferro, 2017 <Ferro-2017>`
   * - 52
-    - DS_OERR
+    - DSS
     - Scoring rule from Equation 16 of :ref:`Dawid and Sebastiani, 1999 <Dawid-Sebastiani-1999>`
-  * - 53
-    - DS_ADD_OERR
-    - Additive observation error scoring rule from Equation 17 of :ref:`Dawid and Sebastiani, 1999 <Dawid-Sebastiani-1999>`
-  * - 54
-    - DS_MULT_OERR
-    - Multiplicative observation error scoring rule from Equation 18 of :ref:`Dawid and Sebastiani, 1999 <Dawid-Sebastiani-1999>`
 
 .. _table_ES_header_info_es_out_RPS:
 

internal/test_unit/hdr/met_12_0.hdr

@@ -19,7 +19,7 @@ PJC           : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_L
 PRC           : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_THRESH _VAR_
 PSTD          : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_THRESH BASER BASER_NCL BASER_NCU RELIABILITY RESOLUTION UNCERTAINTY ROC_AUC BRIER BRIER_NCL BRIER_NCU BRIERCL BRIERCL_NCL BRIERCL_NCU BSS BSS_SMPL _VAR_
 ECLV          : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL BASE N_PTS _VAR_
-ECNT          : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_ENS CRPS CRPSS IGN ME RMSE SPREAD ME_OERR RMSE_OERR SPREAD_OERR SPREAD_PLUS_OERR CRPSCL CRPS_EMP CRPSCL_EMP CRPSS_EMP CRPS_EMP_FAIR SPREAD_MD MAE MAE_OERR BIAS_RATIO N_GE_OBS ME_GE_OBS N_LT_OBS ME_LT_OBS IGN_CONV_OERR IGN_CORR_OERR DS_OERR DS_ADD_OERR DS_MULT_OERR
+ECNT          : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_ENS CRPS CRPSS IGN ME RMSE SPREAD ME_OERR RMSE_OERR SPREAD_OERR SPREAD_PLUS_OERR CRPSCL CRPS_EMP CRPSCL_EMP CRPSS_EMP CRPS_EMP_FAIR SPREAD_MD MAE MAE_OERR BIAS_RATIO N_GE_OBS ME_GE_OBS N_LT_OBS ME_LT_OBS IGN_CONV_OERR IGN_CORR_OERR DSS
 RPS           : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL N_PROB RPS_REL RPS_RES RPS_UNC RPS RPSS RPSS_SMPL RPS_COMP
 RHIST         : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL CRPS IGN N_RANK CRPSS SPREAD _VAR_
 PHIST         : VERSION MODEL DESC FCST_LEAD FCST_VALID_BEG FCST_VALID_END OBS_LEAD OBS_VALID_BEG OBS_VALID_END FCST_VAR FCST_UNITS FCST_LEV OBS_VAR OBS_UNITS OBS_LEV OBTYPE VX_MASK INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA LINE_TYPE TOTAL BIN_SIZE N_BIN _VAR_

src/basic/vx_util/stat_column_defs.h

@@ -277,7 +277,7 @@ static const char * ecnt_columns [] = {
    "MAE",         "MAE_OERR",      "BIAS_RATIO",
    "N_GE_OBS",    "ME_GE_OBS",     "N_LT_OBS",
    "ME_LT_OBS",   "IGN_CONV_OERR", "IGN_CORR_OERR",
-   "DS_OERR",     "DS_ADD_OERR",   "DS_MULT_OERR"
+   "DSS"
 };
 
 static const char * rps_columns [] = {

src/libcode/vx_stat_out/stat_columns.cc

@@ -4276,7 +4276,7 @@ void write_ecnt_cols(const ECNTInfo &ecnt_info,
    //    MAE,       MAE_OERR,      BIAS_RATIO,
    //    N_GE_OBS,  ME_GE_OBS,     N_LT_OBS,
    //    ME_LT_OBS, IGN_CONV_OERR, IGN_CORR_OERR,
-   //    DS_OERR,   DS_ADD_OERR,   DS_MULT_OERR
+   //    DSS
    //
    at.set_entry(r, c+0,  // Total Number of Pairs
       ecnt_info.n_pair);
@@ -4359,14 +4359,8 @@ void write_ecnt_cols(const ECNTInfo &ecnt_info,
    at.set_entry(r, c+26,  // Ignorance Score, observation error corrected
       ecnt_info.ign_corr_oerr);
 
-   at.set_entry(r, c+27,  // Dawid-Sebastiani
-      ecnt_info.ds_oerr);
-
-   at.set_entry(r, c+28,  // Dawid-Sebastiani additive error
-      ecnt_info.ds_add_oerr);
-
-   at.set_entry(r, c+29,  // Dawid-Sebastiani multiplicative error
-      ecnt_info.ds_mult_oerr);
+   at.set_entry(r, c+27,  // Dawid-Sebastiani Score
+      ecnt_info.dss);
 
    return;
 }

src/libcode/vx_statistics/ens_stats.cc

@@ -186,7 +186,7 @@ void ECNTInfo::clear() {
    spread_plus_oerr = bad_data_double;
 
    ign_conv_oerr = ign_corr_oerr = bad_data_double;
-   ds_oerr       = ds_add_oerr   = ds_mult_oerr = bad_data_double;
+   dss           = bad_data_double;
 
    n_ge_obs  = n_lt_obs  = 0;
    me_ge_obs = me_lt_obs = bias_ratio = bad_data_double;
@@ -227,9 +227,7 @@ void ECNTInfo::assign(const ECNTInfo &c) {
 
    ign_conv_oerr    = c.ign_conv_oerr;
    ign_corr_oerr    = c.ign_corr_oerr;
-   ds_oerr          = c.ds_oerr;
-   ds_add_oerr      = c.ds_add_oerr;
-   ds_mult_oerr     = c.ds_mult_oerr;
+   dss              = c.dss;
 
    n_ge_obs         = c.n_ge_obs;
    n_lt_obs         = c.n_lt_obs;
@@ -373,9 +371,7 @@ void ECNTInfo::set(const PairDataEnsemble &pd) {
    // Compute log scores with observational uncertainty
    ign_conv_oerr = pd.ign_conv_oerr_na.wmean(pd.wgt_na);
    ign_corr_oerr = pd.ign_corr_oerr_na.wmean(pd.wgt_na);
-   ds_oerr       = pd.ds_oerr_na.wmean(pd.wgt_na);
-   ds_add_oerr   = pd.ds_add_oerr_na.wmean(pd.wgt_na);
-   ds_mult_oerr  = pd.ds_mult_oerr_na.wmean(pd.wgt_na);
+   dss           = pd.dss_na.wmean(pd.wgt_na);
 
    // Compute bias ratio terms 
    n_ge_obs  = nint(pd.n_ge_obs_na.sum());

src/libcode/vx_statistics/ens_stats.h

@@ -83,9 +83,8 @@ class ECNTInfo {
       double spread_plus_oerr;
 
       // Log scores that incorporate observational uncertainty
-      // as advised in Ferro (2017)
-      double ign_conv_oerr, ign_corr_oerr;
-      double ds_oerr, ds_add_oerr, ds_mult_oerr;
+      // and Dawid-Sebastiani score, as advised in Ferro (2017)
+      double ign_conv_oerr, ign_corr_oerr, dss;
 
       // Bias ratio information
       int n_ge_obs, n_lt_obs;

src/libcode/vx_statistics/pair_data_ensemble.cc

@@ -110,9 +110,7 @@ void PairDataEnsemble::clear() {
 
    ign_conv_oerr_na.clear();
    ign_corr_oerr_na.clear();
-   ds_oerr_na.clear();
-   ds_add_oerr_na.clear();
-   ds_mult_oerr_na.clear();
+   dss_na.clear();
 
    n_ge_obs_na.clear();
    me_ge_obs_na.clear();
@@ -187,9 +185,7 @@ void PairDataEnsemble::extend(int n) {
    pit_na.extend             (n);
    ign_conv_oerr_na.extend   (n);
    ign_corr_oerr_na.extend   (n);
-   ds_oerr_na.extend         (n);
-   ds_add_oerr_na.extend     (n);
-   ds_mult_oerr_na.extend    (n);
+   dss_na.extend             (n);
    n_ge_obs_na.extend        (n);
    me_ge_obs_na.extend       (n);
    n_lt_obs_na.extend        (n);
@@ -259,9 +255,7 @@ void PairDataEnsemble::assign(const PairDataEnsemble &pd) {
 
    ign_conv_oerr_na = pd.ign_conv_oerr_na;
    ign_corr_oerr_na = pd.ign_corr_oerr_na;
-   ds_oerr_na       = pd.ds_oerr_na;
-   ds_add_oerr_na   = pd.ds_add_oerr_na;
-   ds_mult_oerr_na  = pd.ds_mult_oerr_na;
+   dss_na           = pd.dss_na;
 
    n_ge_obs_na    = pd.n_ge_obs_na;
    me_ge_obs_na   = pd.me_ge_obs_na;
@@ -470,9 +464,7 @@ void PairDataEnsemble::compute_pair_vals(const gsl_rng *rng_ptr) {
          pit_na.add(bad_data_double);
          ign_conv_oerr_na.add(bad_data_double);
          ign_corr_oerr_na.add(bad_data_double);
-         ds_oerr_na.add(bad_data_double);
-         ds_add_oerr_na.add(bad_data_double);
-         ds_mult_oerr_na.add(bad_data_double);
+         dss_na.add(bad_data_double);
          n_ge_obs_na.add(bad_data_double);
          me_ge_obs_na.add(bad_data_double);
          n_lt_obs_na.add(bad_data_double);
@@ -485,48 +477,42 @@ void PairDataEnsemble::compute_pair_vals(const gsl_rng *rng_ptr) {
          var_unperturbed = compute_variance(esum_na[i], esumsq_na[i], esumn_na[i]);
          var_na.add(var_unperturbed);
 
+         double emn_unperturbed = compute_mean(esum_na[i], esumn_na[i]);
+         double esd_unperturbed = compute_stdev(esum_na[i], esumsq_na[i], esumn_na[i]);
+
+         // Compute the Dawid Sebastiani scores
+         dss_na.add(
+            compute_dawid_sebastiani(
+               emn_unperturbed, esd_unperturbed, o_na[i]));
+
          // Process the observation error information
          ObsErrorEntry * e = (has_obs_error() ? obs_error_entry[i] : 0);
          if(e) {
 
-            double emn_unperturbed = compute_mean(esum_na[i], esumn_na[i]);
-            double esd_unperturbed = compute_stdev(esum_na[i], esumsq_na[i], esumn_na[i]);
-            double v_conv, v_corr;
-
             // Compute the observation error log scores
+            double v_conv, v_corr;
             compute_obs_error_log_scores(
-               emn_unperturbed, esd_unperturbed, o_na[i], e->variance(),
+               emn_unperturbed, esd_unperturbed,
+               o_na[i], e->variance(),
                v_conv, v_corr);
             ign_conv_oerr_na.add(v_conv);
             ign_corr_oerr_na.add(v_corr);
 
-            // Compute the Dawid Sebastiani scores
-            double v_ds, v_ds_add, v_ds_mult;
-            compute_dawid_sebastiani(
-               emn_unperturbed, esd_unperturbed, o_na[i], e->variance(),
-               e->bias_scale, e->bias_offset,
-               v_ds, v_ds_add, v_ds_mult);
-            ds_oerr_na.add(v_ds);
-            ds_add_oerr_na.add(v_ds_add);
-            ds_mult_oerr_na.add(v_ds_mult);
-
             // Compute perturbed ensemble mean and variance
             // Exclude the control member from the variance
             mn_oerr_na.add(cur_ens.mean());
             var_oerr_na.add(cur_ens.variance(ctrl_index));
 
             // Compute the variance plus observation error variance.
-            var_plus_oerr_na.add(var_unperturbed +
-                                 dist_var(e->dist_type,
-                                          e->dist_parm[0], e->dist_parm[1]));
+            var_plus_oerr_na.add(
+               var_unperturbed +
+               dist_var(e->dist_type,
+                        e->dist_parm[0], e->dist_parm[1]));
          }
          // If no observation error specified, store bad data values
          else {
             ign_conv_oerr_na.add(bad_data_double);
             ign_corr_oerr_na.add(bad_data_double);
-            ds_oerr_na.add(bad_data_double);
-            ds_add_oerr_na.add(bad_data_double);
-            ds_mult_oerr_na.add(bad_data_double);
             mn_oerr_na.add(bad_data_double);
             var_oerr_na.add(bad_data_double);
             var_plus_oerr_na.add(bad_data_double);
@@ -918,8 +904,7 @@ PairDataEnsemble PairDataEnsemble::subset_pairs_obs_thresh(const SingleThresh &o
       //   crps_emp_na, crps_emp_fair_na, spread_md_na,
       //   crpscl_emp_na, crps_gaus_na, crpscl_gaus_na,
       //   ign_na, pit_na,
-      //   ign_conv_oerr, ign_corr_oerr,
-      //   ds_oerr, ds_add_oerr, ds_mult_oerr,
+      //   ign_conv_oerr, ign_corr_oerr, dss,
       //   n_gt_obs_na, me_gt_obs_na, n_lt_obs_na, me_lt_obs_na,
       //   var_na, var_oerr_na, var_plus_oerr_na,
       //   mn_na, mn_oerr_na, e_na
@@ -945,9 +930,7 @@ PairDataEnsemble PairDataEnsemble::subset_pairs_obs_thresh(const SingleThresh &o
       pd.pit_na.add(pit_na[i]);
       pd.ign_conv_oerr_na.add(ign_conv_oerr_na[i]);
       pd.ign_corr_oerr_na.add(ign_corr_oerr_na[i]);
-      pd.ds_oerr_na.add(ds_oerr_na[i]);
-      pd.ds_add_oerr_na.add(ds_add_oerr_na[i]);
-      pd.ds_mult_oerr_na.add(ds_mult_oerr_na[i]);
+      pd.dss_na.add(dss_na[i]);
       pd.n_ge_obs_na.add(n_ge_obs_na[i]);
       pd.me_ge_obs_na.add(me_ge_obs_na[i]);
       pd.n_lt_obs_na.add(n_lt_obs_na[i]);
@@ -2220,79 +2203,36 @@ void compute_obs_error_log_scores(double emn, double esd,
 
 ////////////////////////////////////////////////////////////////////////
 
-void compute_dawid_sebastiani(double emn, double esd,
-                              double obs, double oerr_var,
-                              double oerr_mbias, double oerr_abias,
-                              double &ds_oerr, double &ds_add_oerr,
-                              double &ds_mult_oerr) {
+double compute_dawid_sebastiani(double emn, double esd, double obs) {
 
    const char *method_name = "compute_dawid_sebastiani() -> ";
 
+   double v = bad_data_double;
+
    // Compute the Dawid-Sebastiani scoring rules in
    // Ferro (2017, Eqs 17 and 18) doi:10.1002/qj.3115
    // Dawid and Sebastiani (1999) doi:10.1214/aos/1018031101
    // These are the recommended scores in Ferro (2017).
 
-   // Equation 16 (no obs uncertainty)
-   if(is_bad_data(emn) ||
-      is_bad_data(esd) ||
-      is_bad_data(obs) ||
-      is_eq(esd, 0.0)) {
-      ds_oerr = bad_data_double;
-   }
-   else {
-      ds_oerr = log(esd) +
-                (obs - emn) * (obs - emn) /
-                (2.0 * esd * esd);
-   }
+   // Check for bad data
+   if(!is_bad_data(emn) && !is_bad_data(esd) &&
+      !is_bad_data(obs) && !is_eq(esd, 0.0)) {
 
-   // Equations 17 and 18
-   if(is_bad_data(emn) ||
-      is_bad_data(esd) ||
-      is_bad_data(obs) ||
-      is_bad_data(oerr_var)) {
-      ds_add_oerr = ds_mult_oerr = bad_data_double;
-   }
-   else {
-
-      // Default additive and multiplicative biases to 0 and 1
-      double a = (is_bad_data(oerr_abias) ? 0.0 : oerr_abias);
-      double b = (is_bad_data(oerr_mbias) ? 1.0 : oerr_mbias);
-
-      double b2s2 = 2.0 * b * b * esd * esd;
-      double ov2  = oerr_var * oerr_var;
-
-      if(is_eq(b2s2, 0.0)) {
-         ds_add_oerr = ds_mult_oerr = bad_data_double;
-      }
-      else {
-         ds_add_oerr = log(esd) +
-                       ((obs - a - b * emn) *
-                        (obs - a - b * emn) - ov2) /
-                       b2s2;
-
-         ds_mult_oerr = log(esd) +
-                        ((obs - b * emn) *
-                         (obs - b * emn) - obs * obs * ov2 /
-                         (b * b + ov2)) /
-                        b2s2;
-      }
+      // Equation 16
+      v = log(esd) +
+          (obs - emn) * (obs - emn) /
+          (2.0 * esd * esd);
    }
 
    if(mlog.verbosity_level() >= 10) {
       mlog << Debug(10) << method_name
-           << "inputs (emn = " << emn
+           << "for input emn = " << emn
            << ", esd = " << esd
            << ", obs = " << obs
-           << ", oerr_var = " << oerr_var
-           << ", oerr_mbias = " << oerr_mbias
-           << ", oerr_abias = " << oerr_abias
-           << ") and outputs (ds_oerr = " << ds_oerr
-           << ", ds_add_oerr = " << ds_add_oerr
-           << ", ds_mult_oerr = " << ds_mult_oerr << ")\n";
+           << ", output dss = " << v << "\n";
    }
 
-   return;
+   return(v);
 }
 
 ////////////////////////////////////////////////////////////////////////

src/libcode/vx_statistics/pair_data_ensemble.h

@@ -93,9 +93,7 @@ class PairDataEnsemble : public PairBase {
 
       NumArray   ign_conv_oerr_na; // Error convolved log score [n_obs]
       NumArray   ign_corr_oerr_na; // Error corrected log score [n_obs]
-      NumArray   ds_oerr_na;       // Dawid Sebastiani score [n_obs]
-      NumArray   ds_add_oerr_na;   // Dawid Sebastiani additive error [n_obs]
-      NumArray   ds_mult_oerr_na;  // Dawid Sebastiani multiplicative error [n_obs]
+      NumArray   dss_na;           // Dawid Sebastiani score [n_obs]
 
       NumArray   n_ge_obs_na;      // Number of ensemble memebers >= obs [n_obs]
       NumArray   me_ge_obs_na;     // Mean error of ensemble members >= obs [n_obs]
@@ -334,9 +332,8 @@ extern double compute_bias_ratio(double, double);
 extern void   compute_obs_error_log_scores(
                  double, double, double, double,
                  double &, double &);
-extern void   compute_dawid_sebastiani(
-                 double, double, double, double, double, double,
-                 double &, double &, double &);
+extern double compute_dawid_sebastiani(
+                 double, double, double);
 
 ////////////////////////////////////////////////////////////////////////