diff options
| author | TinWoodman92 <chrhodgden@gmail.com> | 2023-12-16 06:18:44 -0600 |
|---|---|---|
| committer | TinWoodman92 <chrhodgden@gmail.com> | 2023-12-16 06:18:44 -0600 |
| commit | 158b36e51c8c2925d6c8af6e08380369d019378e (patch) | |
| tree | f91fadecf85e982714927eb6ab451c9173e37de0 /alg_move.sql | |
Initial commit
Diffstat (limited to 'alg_move.sql')
| -rw-r--r-- | alg_move.sql | 68 |
1 files changed, 68 insertions, 0 deletions
diff --git a/alg_move.sql b/alg_move.sql new file mode 100644 index 0000000..3ac8cad --- /dev/null +++ b/alg_move.sql @@ -0,0 +1,68 @@ +-- Does expression * NULL return NULL? +-- How about with other arithmetic operators? +-- If so, we can simplify the aggregation to eliminate CASE and/or nested CASE statements +SELECT + move_log.move, + count(move_log.move) AS move_count, + count(CASE + WHEN game_log.victor_name = :plr_name AND game_log.victor_color_int = :plr_color_int + THEN move_log.move ELSE NULL END) AS win_count, + count(CASE + WHEN game_log.vic_chk = 0 + THEN move_log.move ELSE NULL END) AS draw_count, + count(CASE + WHEN game_log.vic_chk = 1 AND (game_log.victor_name <> :plr_name OR game_log.victor_color_int <> :plr_color_int) + THEN move_log.move ELSE NULL END) AS loss_count, + sum(CASE + WHEN game_log.victor_name = :plr_name AND game_log.victor_color_int = :plr_color_int + THEN (1.0 / 1.0) + ELSE NULL END) AS sum_win_factor, + sum(CASE + WHEN game_log.vic_chk = 0 + THEN (0.0 / 1.0) + ELSE NULL END) AS sum_draw_factor, + sum(CASE + WHEN game_log.vic_chk = 1 AND (game_log.victor_name <> :plr_name OR game_log.victor_color_int <> :plr_color_int) + THEN (-1.0 / 1.0) + ELSE NULL END) AS sum_loss_factor, + avg(CASE + WHEN game_log.victor_name = :plr_name AND game_log.victor_color_int = :plr_color_int + THEN (1.0 / 1.0) + ELSE NULL END) AS avg_win_factor, + avg(CASE + WHEN game_log.vic_chk = 0 + THEN (0.0 / 1.0) + ELSE NULL END) AS avg_draw_factor, + avg(CASE + WHEN game_log.vic_chk = 1 AND (game_log.victor_name <> :plr_name OR game_log.victor_color_int <> :plr_color_int) + THEN (-1.0 / 1.0) + ELSE NULL END) AS avg_loss_factor, + sum(CASE + WHEN game_log.victor_name = :plr_name AND game_log.victor_color_int = :plr_color_int + THEN (1.0 / 1.0) + WHEN game_log.vic_chk = 0 + THEN (0.0 / 1.0) + WHEN game_log.vic_chk = 1 AND (game_log.victor_name <> :plr_name OR game_log.victor_color_int <> :plr_color_int) + THEN (-1.0 / 1.0) + ELSE NULL END) AS sum_net_factor, + avg(CASE + WHEN game_log.victor_name = :plr_name AND game_log.victor_color_int = :plr_color_int + THEN (1.0 / 1.0) + WHEN game_log.vic_chk = 0 + THEN (0.0 / 1.0) + WHEN game_log.vic_chk = 1 AND (game_log.victor_name <> :plr_name OR game_log.victor_color_int <> :plr_color_int) + THEN (-1.0 / 1.0) + ELSE NULL END) AS avg_net_factor +FROM + game_log INNER JOIN move_log + ON game_log.game_id = move_log.game_id +WHERE + move_log.board = :fen_board + AND game_log.plr_X_name = :plr_name + AND game_log.plr_X_color_int = :plr_color_int +GROUP BY + move_log.move +ORDER BY + (win_count - loss_count) DESC, + move_log.move +;
\ No newline at end of file |