feat: higher order Taylor expansions (quad/oct)

README.md

@@ -11,6 +11,10 @@
 Include the header in one C file with `POLESITTER_IMPLEMENTATION` defined. \
 Define `PS_MULTITHREADING` as well when using more than one thread. \
 Define `PS_2D` when simulating in two-dimensional space.
+Use #define PS_ORDER 1-3 to pick maximum order of Taylor expansions used in solver. \
+    * (higher = worse performance, better accuracy)
+    * (default = 2)
+
 
 ```c
 #define PS_MULTITHREADING

demo/binary.c

@@ -11,7 +11,7 @@
 #define BLACKHOLE_MASS     1000.0F
 #define ARENA_SIZE         (1024ULL * 1024ULL * 256ULL)
 #define MAX_EXPECTED_SPEED 20
-#define THETA              2.0F // sqrt of 12
+#define THETA              2.0F
 
 float    vx[PARTICLE_COUNT]           = {0};
 float    vy[PARTICLE_COUNT]           = {0};

demo/unary.c

@@ -12,7 +12,7 @@
 #define BLACKHOLE_MASS     1000.0F
 #define ARENA_SIZE         (1024ULL * 1024ULL * 256ULL)
 #define MAX_EXPECTED_SPEED 20
-#define THETA              2.00F // sqrt of 12
+#define THETA              2.0F
 
 float    vx[PARTICLE_COUNT]           = {0};
 float    vy[PARTICLE_COUNT]           = {0};

demo/unary_2d.c

@@ -13,7 +13,7 @@
 #define BLACKHOLE_MASS     7000.0F
 #define ARENA_SIZE         (1024ULL * 1024ULL * 512ULL)
 #define MAX_EXPECTED_SPEED 750
-#define THETA              3.4641F // sqrt of 12
+#define THETA              2.0F
 
 float    vx[PARTICLE_COUNT]           = {0};
 float    vy[PARTICLE_COUNT]           = {0};