We get about 25% performance gain

with the custom lsoda call, avoiding repeated getNativeSymbolInfo calls.
It's just that we should not be calling foreign functions from different
packages, because the may change without notice. Using
getNativeSymbolInfo for "call_lsoda" avoids the CRAN note, and a similar
call could probably be used for "unlock_solver", avoiding the NOTE in
checks for cran, but we should not do this in a CRAN package.

1 files changed, 47 insertions, 12 deletions

diff --git a/vignettes/web_only/saem_benchmarks.html b/vignettes/web_only/saem_benchmarks.html
index 4875bb1b..714dc1ff 100644
--- a/vignettes/web_only/saem_benchmarks.html
+++ b/vignettes/web_only/saem_benchmarks.html
@@ -1599,7 +1599,7 @@ div.tocify {
 
 <h1 class="title toc-ignore">Benchmark timings for saem.mmkin</h1>
 <h4 class="author">Johannes Ranke</h4>
-<h4 class="date">Last change 14 November 2022 (rebuilt 2022-11-14)</h4>
+<h4 class="date">Last change 14 November 2022 (rebuilt 2022-11-15)</h4>
 
 </div>
 
@@ -1781,10 +1781,20 @@ t11 &lt;- system.time(sforb_sfo3_plus_const &lt;- saem(three_met_sep_tc[&quot;SF
 <td align="left">Linux</td>
 <td align="left">1.2.0</td>
 <td align="left">3.2</td>
-<td align="right">2.996</td>
-<td align="right">5.207</td>
-<td align="right">5.317</td>
-<td align="right">5.171</td>
+<td align="right">2.110</td>
+<td align="right">4.632</td>
+<td align="right">4.264</td>
+<td align="right">4.930</td>
+</tr>
+<tr class="even">
+<td align="left">Ryzen 7 1700</td>
+<td align="left">Linux</td>
+<td align="left">1.3.0</td>
+<td align="left">3.2</td>
+<td align="right">2.394</td>
+<td align="right">4.748</td>
+<td align="right">4.883</td>
+<td align="right">4.937</td>
 </tr>
 </tbody>
 </table>
@@ -1808,10 +1818,20 @@ t11 &lt;- system.time(sforb_sfo3_plus_const &lt;- saem(three_met_sep_tc[&quot;SF
 <td align="left">Linux</td>
 <td align="left">1.2.0</td>
 <td align="left">3.2</td>
-<td align="right">5.671</td>
-<td align="right">7.696</td>
-<td align="right">8.166</td>
-<td align="right">8.168</td>
+<td align="right">5.602</td>
+<td align="right">7.373</td>
+<td align="right">7.815</td>
+<td align="right">7.831</td>
+</tr>
+<tr class="even">
+<td align="left">Ryzen 7 1700</td>
+<td align="left">Linux</td>
+<td align="left">1.3.0</td>
+<td align="left">3.2</td>
+<td align="right">5.622</td>
+<td align="right">7.445</td>
+<td align="right">8.297</td>
+<td align="right">7.740</td>
 </tr>
 </tbody>
 </table>
@@ -1836,8 +1856,16 @@ t11 &lt;- system.time(sforb_sfo3_plus_const &lt;- saem(three_met_sep_tc[&quot;SF
 <td align="left">Linux</td>
 <td align="left">1.2.0</td>
 <td align="left">3.2</td>
-<td align="right">24.883</td>
-<td align="right">818.157</td>
+<td align="right">24.014</td>
+<td align="right">749.699</td>
+</tr>
+<tr class="even">
+<td align="left">Ryzen 7 1700</td>
+<td align="left">Linux</td>
+<td align="left">1.3.0</td>
+<td align="left">3.2</td>
+<td align="right">24.480</td>
+<td align="right">519.087</td>
 </tr>
 </tbody>
 </table>
@@ -1861,7 +1889,14 @@ t11 &lt;- system.time(sforb_sfo3_plus_const &lt;- saem(three_met_sep_tc[&quot;SF
 <td align="left">Linux</td>
 <td align="left">1.2.0</td>
 <td align="left">3.2</td>
-<td align="right">1355.036</td>
+<td align="right">1249.834</td>
+</tr>
+<tr class="even">
+<td align="left">Ryzen 7 1700</td>
+<td align="left">Linux</td>
+<td align="left">1.3.0</td>
+<td align="left">3.2</td>
+<td align="right">944.471</td>
 </tr>
 </tbody>
 </table>