numo-narray-alt 0.9.4 → 0.9.5

data/ext/numo/narray/src/t_sfloat.c

@@ -237,8 +237,8 @@ static void iter_sfloat_store_bit(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_bit(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_bit, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_bit, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -286,8 +286,8 @@ static void iter_sfloat_store_dfloat(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_dfloat(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_dfloat, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_dfloat, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -335,8 +335,8 @@ static void iter_sfloat_store_sfloat(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_sfloat(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_sfloat, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_sfloat, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -384,8 +384,8 @@ static void iter_sfloat_store_int64(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_int64(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_int64, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_int64, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -433,8 +433,8 @@ static void iter_sfloat_store_int32(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_int32(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_int32, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_int32, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -482,8 +482,8 @@ static void iter_sfloat_store_int16(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_int16(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_int16, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_int16, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -531,8 +531,8 @@ static void iter_sfloat_store_int8(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_int8(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_int8, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_int8, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -580,8 +580,8 @@ static void iter_sfloat_store_uint64(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_uint64(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_uint64, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_uint64, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -629,8 +629,8 @@ static void iter_sfloat_store_uint32(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_uint32(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_uint32, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_uint32, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -678,8 +678,8 @@ static void iter_sfloat_store_uint16(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_uint16(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_uint16, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_uint16, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -727,8 +727,8 @@ static void iter_sfloat_store_uint8(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_uint8(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_uint8, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_uint8, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -776,8 +776,8 @@ static void iter_sfloat_store_robject(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_store_robject(VALUE self, VALUE obj) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_robject, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_robject, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, obj);
   return self;
@@ -885,8 +885,8 @@ loop_end:
 }
 
 static VALUE sfloat_store_array(VALUE self, VALUE rary) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {rb_cArray, 0}};
-  ndfunc_t ndf = {iter_sfloat_store_array, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { rb_cArray, 0 } };
+  ndfunc_t ndf = { iter_sfloat_store_array, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop_store_rarray(&ndf, self, rary);
   return self;
@@ -981,8 +981,10 @@ static VALUE sfloat_store(VALUE self, VALUE obj) {
     }
   }
 
-  rb_raise(nary_eCastError, "unknown conversion from %s to %s", rb_class2name(rb_obj_class(obj)),
-           rb_class2name(rb_obj_class(self)));
+  rb_raise(
+    nary_eCastError, "unknown conversion from %s to %s", rb_class2name(rb_obj_class(obj)),
+    rb_class2name(rb_obj_class(self))
+  );
 
   return self;
 }
@@ -1076,7 +1078,10 @@ static dtype sfloat_extract_data(VALUE obj) {
       return sfloat_extract_data(r);
     }
 
-    rb_raise(nary_eCastError, "unknown conversion from %s to %s", rb_class2name(rb_obj_class(obj)), rb_class2name(cT));
+    rb_raise(
+      nary_eCastError, "unknown conversion from %s to %s", rb_class2name(rb_obj_class(obj)),
+      rb_class2name(cT)
+    );
   }
   if (TYPE(obj) == T_ARRAY) {
     if (RARRAY_LEN(obj) != 1) {
@@ -1145,8 +1150,8 @@ static VALUE sfloat_s_cast(VALUE type, VALUE obj) {
 /*
   Multi-dimensional element reference.
   @overload [](dim0,...,dimL)
-  @param [Numeric,Range,Array,Numo::Int32,Numo::Int64,Numo::Bit,TrueClass,FalseClass,Symbol] dim0,...,dimL  multi-dimensional
-  indices.
+  @param [Numeric,Range,Array,Numo::Int32,Numo::Int64,Numo::Bit,TrueClass,FalseClass,Symbol]
+  dim0,...,dimL  multi-dimensional indices.
   @return [Numeric,Numo::SFloat] an element or NArray view.
   @see Numo::NArray#[]
   @see #[]=
@@ -1168,8 +1173,8 @@ static VALUE sfloat_aref(int argc, VALUE* argv, VALUE self) {
 /*
   Multi-dimensional element assignment.
   @overload []=(dim0,...,dimL,val)
-  @param [Numeric,Range,Array,Numo::Int32,Numo::Int64,Numo::Bit,TrueClass,FalseClass,Symbol] dim0,...,dimL  multi-dimensional
-  indices.
+  @param [Numeric,Range,Array,Numo::Int32,Numo::Int64,Numo::Bit,TrueClass,FalseClass,Symbol]
+  dim0,...,dimL  multi-dimensional indices.
   @param [Numeric,Numo::NArray,Array] val  Value(s) to be set to self.
   @return [Numeric,Numo::NArray,Array] returns `val` (last argument).
   @see Numo::NArray#[]=
@@ -1240,9 +1245,9 @@ static void iter_sfloat_to_a(na_loop_t* const lp) {
   @return [Array]
 */
 static VALUE sfloat_to_a(VALUE self) {
-  ndfunc_arg_in_t ain[3] = {{Qnil, 0}, {sym_loop_opt}, {sym_option}};
-  ndfunc_arg_out_t aout[1] = {{rb_cArray, 0}}; // dummy?
-  ndfunc_t ndf = {iter_sfloat_to_a, FULL_LOOP_NIP, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { Qnil, 0 }, { sym_loop_opt }, { sym_option } };
+  ndfunc_arg_out_t aout[1] = { { rb_cArray, 0 } }; // dummy?
+  ndfunc_t ndf = { iter_sfloat_to_a, FULL_LOOP_NIP, 3, 1, ain, aout };
   return na_ndloop_cast_narray_to_rarray(&ndf, self, Qnil);
 }
 
@@ -1274,8 +1279,8 @@ static void iter_sfloat_fill(na_loop_t* const lp) {
   @return [Numo::SFloat] self.
 */
 static VALUE sfloat_fill(VALUE self, VALUE val) {
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {sym_option}};
-  ndfunc_t ndf = {iter_sfloat_fill, FULL_LOOP, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { sym_option } };
+  ndfunc_t ndf = { iter_sfloat_fill, FULL_LOOP, 2, 0, ain, 0 };
 
   na_ndloop(&ndf, 2, self, val);
   return self;
@@ -1330,9 +1335,9 @@ static void iter_sfloat_format(na_loop_t* const lp) {
 static VALUE sfloat_format(int argc, VALUE* argv, VALUE self) {
   VALUE fmt = Qnil;
 
-  ndfunc_arg_in_t ain[2] = {{Qnil, 0}, {sym_option}};
-  ndfunc_arg_out_t aout[1] = {{numo_cRObject, 0}};
-  ndfunc_t ndf = {iter_sfloat_format, FULL_LOOP_NIP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { Qnil, 0 }, { sym_option } };
+  ndfunc_arg_out_t aout[1] = { { numo_cRObject, 0 } };
+  ndfunc_t ndf = { iter_sfloat_format, FULL_LOOP_NIP, 2, 1, ain, aout };
 
   rb_scan_args(argc, argv, "01", &fmt);
   return na_ndloop(&ndf, 2, self, fmt);
@@ -1376,9 +1381,9 @@ static void iter_sfloat_format_to_a(na_loop_t* const lp) {
 */
 static VALUE sfloat_format_to_a(int argc, VALUE* argv, VALUE self) {
   VALUE fmt = Qnil;
-  ndfunc_arg_in_t ain[3] = {{Qnil, 0}, {sym_loop_opt}, {sym_option}};
-  ndfunc_arg_out_t aout[1] = {{rb_cArray, 0}}; // dummy?
-  ndfunc_t ndf = {iter_sfloat_format_to_a, FULL_LOOP_NIP, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { Qnil, 0 }, { sym_loop_opt }, { sym_option } };
+  ndfunc_arg_out_t aout[1] = { { rb_cArray, 0 } }; // dummy?
+  ndfunc_t ndf = { iter_sfloat_format_to_a, FULL_LOOP_NIP, 3, 1, ain, aout };
 
   rb_scan_args(argc, argv, "01", &fmt);
   return na_ndloop_cast_narray_to_rarray(&ndf, self, fmt);
@@ -1432,8 +1437,8 @@ static void iter_sfloat_each(na_loop_t* const lp) {
   @see #map
 */
 static VALUE sfloat_each(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_each, FULL_LOOP_NIP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_each, FULL_LOOP_NIP, 1, 0, ain, 0 };
 
   na_ndloop(&ndf, 1, self);
   return self;
@@ -1507,9 +1512,9 @@ static void iter_sfloat_map(na_loop_t* const lp) {
   @return [Numo::SFloat] map of self.
 */
 static VALUE sfloat_map(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_map, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_map, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -1571,8 +1576,8 @@ static void iter_sfloat_each_with_index(na_loop_t* const lp) {
   @see #map_with_index
 */
 static VALUE sfloat_each_with_index(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{Qnil, 0}};
-  ndfunc_t ndf = {iter_sfloat_each_with_index, FULL_LOOP_NIP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { Qnil, 0 } };
+  ndfunc_t ndf = { iter_sfloat_each_with_index, FULL_LOOP_NIP, 1, 0, ain, 0 };
 
   na_ndloop_with_index(&ndf, 1, self);
   return self;
@@ -1662,9 +1667,9 @@ static void iter_sfloat_map_with_index(na_loop_t* const lp) {
   @see #each_with_index
 */
 static VALUE sfloat_map_with_index(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{Qnil, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_map_with_index, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { Qnil, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_map_with_index, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop_with_index(&ndf, 1, self);
 }
@@ -1716,14 +1721,14 @@ static void iter_sfloat_abs(na_loop_t* const lp) {
   @return [Numo::SFloat] abs of self.
 */
 static VALUE sfloat_abs(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cRT, 0}};
-  ndfunc_t ndf = {iter_sfloat_abs, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cRT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_abs, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_add(na_loop_t* const lp) {
@@ -1748,14 +1753,20 @@ static void iter_sfloat_add(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
 #ifdef __SSE2__
       // Check number of elements. & Check same alignment.
-      if ((n >= num_pack) && is_same_aligned3(&((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+      if ((n >= num_pack) &&
+          is_same_aligned3(
+            &((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE
+          )) {
         // Calculate up to the position just before the start of SIMD computation.
-        cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+        cnt = get_count_of_elements_not_aligned_to_simd_size(
+          &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+        );
 #endif
         if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -1817,7 +1828,8 @@ static void iter_sfloat_add(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -1828,9 +1840,12 @@ static void iter_sfloat_add(na_loop_t* const lp) {
           b = _mm_load1_ps(&((dtype*)p2)[0]);
 
           // Check number of elements. & Check same alignment.
-          if ((n >= num_pack) && is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+          if ((n >= num_pack) &&
+              is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
             // Calculate up to the position just before the start of SIMD computation.
-            cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+            cnt = get_count_of_elements_not_aligned_to_simd_size(
+              &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+            );
 #endif
             if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -1926,9 +1941,9 @@ static void iter_sfloat_add(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_add_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_add, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_add, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -1952,7 +1967,7 @@ static VALUE sfloat_add(VALUE self, VALUE other) {
   }
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_sub(na_loop_t* const lp) {
@@ -1977,14 +1992,20 @@ static void iter_sfloat_sub(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
 #ifdef __SSE2__
       // Check number of elements. & Check same alignment.
-      if ((n >= num_pack) && is_same_aligned3(&((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+      if ((n >= num_pack) &&
+          is_same_aligned3(
+            &((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE
+          )) {
         // Calculate up to the position just before the start of SIMD computation.
-        cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+        cnt = get_count_of_elements_not_aligned_to_simd_size(
+          &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+        );
 #endif
         if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2046,7 +2067,8 @@ static void iter_sfloat_sub(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -2057,9 +2079,12 @@ static void iter_sfloat_sub(na_loop_t* const lp) {
           b = _mm_load1_ps(&((dtype*)p2)[0]);
 
           // Check number of elements. & Check same alignment.
-          if ((n >= num_pack) && is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+          if ((n >= num_pack) &&
+              is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
             // Calculate up to the position just before the start of SIMD computation.
-            cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+            cnt = get_count_of_elements_not_aligned_to_simd_size(
+              &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+            );
 #endif
             if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2155,9 +2180,9 @@ static void iter_sfloat_sub(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_sub_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_sub, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_sub, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -2181,7 +2206,7 @@ static VALUE sfloat_sub(VALUE self, VALUE other) {
   }
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_mul(na_loop_t* const lp) {
@@ -2206,14 +2231,20 @@ static void iter_sfloat_mul(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
 #ifdef __SSE2__
       // Check number of elements. & Check same alignment.
-      if ((n >= num_pack) && is_same_aligned3(&((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+      if ((n >= num_pack) &&
+          is_same_aligned3(
+            &((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE
+          )) {
         // Calculate up to the position just before the start of SIMD computation.
-        cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+        cnt = get_count_of_elements_not_aligned_to_simd_size(
+          &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+        );
 #endif
         if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2275,7 +2306,8 @@ static void iter_sfloat_mul(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -2286,9 +2318,12 @@ static void iter_sfloat_mul(na_loop_t* const lp) {
           b = _mm_load1_ps(&((dtype*)p2)[0]);
 
           // Check number of elements. & Check same alignment.
-          if ((n >= num_pack) && is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+          if ((n >= num_pack) &&
+              is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
             // Calculate up to the position just before the start of SIMD computation.
-            cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+            cnt = get_count_of_elements_not_aligned_to_simd_size(
+              &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+            );
 #endif
             if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2384,9 +2419,9 @@ static void iter_sfloat_mul(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_mul_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_mul, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_mul, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -2410,7 +2445,7 @@ static VALUE sfloat_mul(VALUE self, VALUE other) {
   }
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_div(na_loop_t* const lp) {
@@ -2435,14 +2470,20 @@ static void iter_sfloat_div(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
 #ifdef __SSE2__
       // Check number of elements. & Check same alignment.
-      if ((n >= num_pack) && is_same_aligned3(&((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+      if ((n >= num_pack) &&
+          is_same_aligned3(
+            &((dtype*)p1)[i], &((dtype*)p2)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE
+          )) {
         // Calculate up to the position just before the start of SIMD computation.
-        cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+        cnt = get_count_of_elements_not_aligned_to_simd_size(
+          &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+        );
 #endif
         if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2504,7 +2545,8 @@ static void iter_sfloat_div(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -2515,9 +2557,12 @@ static void iter_sfloat_div(na_loop_t* const lp) {
           b = _mm_load1_ps(&((dtype*)p2)[0]);
 
           // Check number of elements. & Check same alignment.
-          if ((n >= num_pack) && is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
+          if ((n >= num_pack) &&
+              is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p3)[i], SIMD_ALIGNMENT_SIZE)) {
             // Calculate up to the position just before the start of SIMD computation.
-            cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+            cnt = get_count_of_elements_not_aligned_to_simd_size(
+              &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+            );
 #endif
             if (p1 == p3) { // inplace case
 #ifdef __SSE2__
@@ -2613,9 +2658,9 @@ static void iter_sfloat_div(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_div_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_div, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_div, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -2639,7 +2684,7 @@ static VALUE sfloat_div(VALUE self, VALUE other) {
   }
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_mod(na_loop_t* const lp) {
@@ -2654,7 +2699,8 @@ static void iter_sfloat_mod(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
       if (p1 == p3) { // inplace case
@@ -2671,7 +2717,8 @@ static void iter_sfloat_mod(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -2729,9 +2776,9 @@ static void iter_sfloat_mod(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_mod_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_mod, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_mod, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -2775,9 +2822,9 @@ static void iter_sfloat_divmod(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_divmod_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_divmod, STRIDE_LOOP, 2, 2, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_divmod, STRIDE_LOOP, 2, 2, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -2836,11 +2883,11 @@ static void iter_sfloat_pow_int32(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_pow_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_in_t ain_i[2] = {{cT, 0}, {numo_cInt32, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_pow, STRIDE_LOOP, 2, 1, ain, aout};
-  ndfunc_t ndf_i = {iter_sfloat_pow_int32, STRIDE_LOOP, 2, 1, ain_i, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_in_t ain_i[2] = { { cT, 0 }, { numo_cInt32, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_pow, STRIDE_LOOP, 2, 1, ain, aout };
+  ndfunc_t ndf_i = { iter_sfloat_pow_int32, STRIDE_LOOP, 2, 1, ain_i, aout };
 
   // fixme : use na.integer?
   if (FIXNUM_P(other) || rb_obj_is_kind_of(other, numo_cInt32)) {
@@ -2936,9 +2983,9 @@ static void iter_sfloat_minus(na_loop_t* const lp) {
   @return [Numo::SFloat] minus of self.
 */
 static VALUE sfloat_minus(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_minus, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_minus, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3011,9 +3058,9 @@ static void iter_sfloat_reciprocal(na_loop_t* const lp) {
   @return [Numo::SFloat] reciprocal of self.
 */
 static VALUE sfloat_reciprocal(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_reciprocal, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_reciprocal, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3086,9 +3133,9 @@ static void iter_sfloat_sign(na_loop_t* const lp) {
   @return [Numo::SFloat] sign of self.
 */
 static VALUE sfloat_sign(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_sign, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_sign, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3161,9 +3208,9 @@ static void iter_sfloat_square(na_loop_t* const lp) {
   @return [Numo::SFloat] square of self.
 */
 static VALUE sfloat_square(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_square, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_square, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3190,9 +3237,9 @@ static void iter_sfloat_eq(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_eq_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_eq, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_eq, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3237,9 +3284,9 @@ static void iter_sfloat_ne(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_ne_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_ne, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_ne, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3284,9 +3331,9 @@ static void iter_sfloat_nearly_eq(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_nearly_eq_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_nearly_eq, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_nearly_eq, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3377,9 +3424,9 @@ static void iter_sfloat_floor(na_loop_t* const lp) {
   @return [Numo::SFloat] floor of self.
 */
 static VALUE sfloat_floor(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_floor, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_floor, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3452,9 +3499,9 @@ static void iter_sfloat_round(na_loop_t* const lp) {
   @return [Numo::SFloat] round of self.
 */
 static VALUE sfloat_round(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_round, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_round, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3527,9 +3574,9 @@ static void iter_sfloat_ceil(na_loop_t* const lp) {
   @return [Numo::SFloat] ceil of self.
 */
 static VALUE sfloat_ceil(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_ceil, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_ceil, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3602,9 +3649,9 @@ static void iter_sfloat_trunc(na_loop_t* const lp) {
   @return [Numo::SFloat] trunc of self.
 */
 static VALUE sfloat_trunc(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_trunc, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_trunc, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3677,14 +3724,14 @@ static void iter_sfloat_rint(na_loop_t* const lp) {
   @return [Numo::SFloat] rint of self.
 */
 static VALUE sfloat_rint(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_rint, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_rint, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
 
-#define check_intdivzero(y)                                                                                                    \
+#define check_intdivzero(y) \
   {}
 
 static void iter_sfloat_copysign(na_loop_t* const lp) {
@@ -3699,7 +3746,8 @@ static void iter_sfloat_copysign(na_loop_t* const lp) {
   INIT_PTR(lp, 2, p3, s3);
 
   //
-  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) && is_aligned(p3, sizeof(dtype))) {
+  if (is_aligned(p1, sizeof(dtype)) && is_aligned(p2, sizeof(dtype)) &&
+      is_aligned(p3, sizeof(dtype))) {
 
     if (s1 == sizeof(dtype) && s2 == sizeof(dtype) && s3 == sizeof(dtype)) {
       if (p1 == p3) { // inplace case
@@ -3716,7 +3764,8 @@ static void iter_sfloat_copysign(na_loop_t* const lp) {
       return;
     }
 
-    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) && is_aligned_step(s3, sizeof(dtype))) {
+    if (is_aligned_step(s1, sizeof(dtype)) && is_aligned_step(s2, sizeof(dtype)) &&
+        is_aligned_step(s3, sizeof(dtype))) {
       //
 
       if (s2 == 0) { // Broadcasting from scalar value.
@@ -3774,9 +3823,9 @@ static void iter_sfloat_copysign(na_loop_t* const lp) {
 #undef check_intdivzero
 
 static VALUE sfloat_copysign_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_copysign, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_copysign, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3835,9 +3884,9 @@ static void iter_sfloat_signbit(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of signbit.
 */
 static VALUE sfloat_signbit(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_signbit, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_signbit, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3865,9 +3914,9 @@ static void iter_sfloat_modf(na_loop_t* const lp) {
   @return [Numo::SFloat] modf of self.
 */
 static VALUE sfloat_modf(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_modf, STRIDE_LOOP, 1, 2, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_modf, STRIDE_LOOP, 1, 2, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -3894,9 +3943,9 @@ static void iter_sfloat_gt(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_gt_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_gt, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_gt, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3941,9 +3990,9 @@ static void iter_sfloat_ge(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_ge_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_ge, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_ge, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -3988,9 +4037,9 @@ static void iter_sfloat_lt(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_lt_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_lt, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_lt, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -4035,9 +4084,9 @@ static void iter_sfloat_le(na_loop_t* const lp) {
 }
 
 static VALUE sfloat_le_self(VALUE self, VALUE other) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_le, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_le, STRIDE_LOOP, 2, 1, ain, aout };
 
   return na_ndloop(&ndf, 2, self, other);
 }
@@ -4156,11 +4205,11 @@ static void iter_sfloat_clip_max(na_loop_t* const lp) {
       # [3, 4, 2, 3, 4, 5, 6, 7, 8, 8]
 */
 static VALUE sfloat_clip(VALUE self, VALUE min, VALUE max) {
-  ndfunc_arg_in_t ain[3] = {{Qnil, 0}, {cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf_min = {iter_sfloat_clip_min, STRIDE_LOOP, 2, 1, ain, aout};
-  ndfunc_t ndf_max = {iter_sfloat_clip_max, STRIDE_LOOP, 2, 1, ain, aout};
-  ndfunc_t ndf_both = {iter_sfloat_clip, STRIDE_LOOP, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { Qnil, 0 }, { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf_min = { iter_sfloat_clip_min, STRIDE_LOOP, 2, 1, ain, aout };
+  ndfunc_t ndf_max = { iter_sfloat_clip_max, STRIDE_LOOP, 2, 1, ain, aout };
+  ndfunc_t ndf_both = { iter_sfloat_clip, STRIDE_LOOP, 3, 1, ain, aout };
 
   if (RTEST(min)) {
     if (RTEST(max)) {
@@ -4212,9 +4261,9 @@ static void iter_sfloat_isnan(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of isnan.
 */
 static VALUE sfloat_isnan(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_isnan, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_isnan, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -4254,9 +4303,9 @@ static void iter_sfloat_isinf(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of isinf.
 */
 static VALUE sfloat_isinf(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_isinf, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_isinf, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -4296,9 +4345,9 @@ static void iter_sfloat_isposinf(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of isposinf.
 */
 static VALUE sfloat_isposinf(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_isposinf, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_isposinf, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -4338,9 +4387,9 @@ static void iter_sfloat_isneginf(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of isneginf.
 */
 static VALUE sfloat_isneginf(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_isneginf, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_isneginf, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -4380,9 +4429,9 @@ static void iter_sfloat_isfinite(na_loop_t* const lp) {
   @return [Numo::Bit] Condition of isfinite.
 */
 static VALUE sfloat_isfinite(VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{numo_cBit, 0}};
-  ndfunc_t ndf = {iter_sfloat_isfinite, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { numo_cBit, 0 } };
+  ndfunc_t ndf = { iter_sfloat_isfinite, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, self);
 }
@@ -4413,16 +4462,18 @@ static void iter_sfloat_sum_nan(na_loop_t* const lp) {
 /*
   sum of self.
   @overload sum(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs sum along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of sum.
 */
 static VALUE sfloat_sum(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_sum, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_sum, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_sum_nan);
 
@@ -4457,16 +4508,18 @@ static void iter_sfloat_prod_nan(na_loop_t* const lp) {
 /*
   prod of self.
   @overload prod(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs prod along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of prod.
 */
 static VALUE sfloat_prod(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_prod, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_prod, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_prod_nan);
 
@@ -4501,16 +4554,18 @@ static void iter_sfloat_mean_nan(na_loop_t* const lp) {
 /*
   mean of self.
   @overload mean(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs mean along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of mean.
 */
 static VALUE sfloat_mean(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_mean, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_mean, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_mean_nan);
 
@@ -4545,16 +4600,18 @@ static void iter_sfloat_stddev_nan(na_loop_t* const lp) {
 /*
   stddev of self.
   @overload stddev(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs stddev along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of stddev.
 */
 static VALUE sfloat_stddev(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cRT, 0}};
-  ndfunc_t ndf = {iter_sfloat_stddev, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cRT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_stddev, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_stddev_nan);
 
@@ -4589,16 +4646,18 @@ static void iter_sfloat_var_nan(na_loop_t* const lp) {
 /*
   var of self.
   @overload var(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs var along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of var.
 */
 static VALUE sfloat_var(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cRT, 0}};
-  ndfunc_t ndf = {iter_sfloat_var, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cRT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_var, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_var_nan);
 
@@ -4633,16 +4692,18 @@ static void iter_sfloat_rms_nan(na_loop_t* const lp) {
 /*
   rms of self.
   @overload rms(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs rms along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of rms.
 */
 static VALUE sfloat_rms(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cRT, 0}};
-  ndfunc_t ndf = {iter_sfloat_rms, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cRT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_rms, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_rms_nan);
 
@@ -4677,16 +4738,18 @@ static void iter_sfloat_min_nan(na_loop_t* const lp) {
 /*
   min of self.
   @overload min(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs min along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of min.
 */
 static VALUE sfloat_min(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_min, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_min, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_min_nan);
 
@@ -4721,16 +4784,18 @@ static void iter_sfloat_max_nan(na_loop_t* const lp) {
 /*
   max of self.
   @overload max(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs max along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of max.
 */
 static VALUE sfloat_max(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_max, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_max, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_max_nan);
 
@@ -4765,16 +4830,18 @@ static void iter_sfloat_ptp_nan(na_loop_t* const lp) {
 /*
   ptp of self.
   @overload ptp(axis:nil, keepdims:false, nan:false)
-  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or, return NaN for min/max etc).
+  @param [TrueClass] nan  If true, apply NaN-aware algorithm (avoid NaN for sum/mean etc, or,
+  return NaN for min/max etc).
   @param [Numeric,Array,Range] axis  Performs ptp along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns result of ptp.
 */
 static VALUE sfloat_ptp(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_ptp, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_ptp, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_ptp_nan);
 
@@ -4855,7 +4922,8 @@ static void iter_sfloat_max_index_index32_nan(na_loop_t* const lp) {
   Index of the maximum value.
   @overload max_index(axis:nil, nan:false)
   @param [TrueClass] nan  If true, apply NaN-aware algorithm (return NaN posision if exist).
-  @param [Numeric,Array,Range] axis  Finds maximum values along the axis and returns **flat 1-d indices**.
+  @param [Numeric,Array,Range] axis  Finds maximum values along the axis and returns **flat 1-d
+  indices**.
   @return [Integer,Numo::Int] returns result indices.
   @see #argmax
   @see #max
@@ -4873,9 +4941,9 @@ static void iter_sfloat_max_index_index32_nan(na_loop_t* const lp) {
 static VALUE sfloat_max_index(int argc, VALUE* argv, VALUE self) {
   narray_t* na;
   VALUE idx, reduce;
-  ndfunc_arg_in_t ain[3] = {{Qnil, 0}, {Qnil, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{0, 0, 0}};
-  ndfunc_t ndf = {0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { Qnil, 0 }, { Qnil, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { 0, 0, 0 } };
+  ndfunc_t ndf = { 0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 3, 1, ain, aout };
 
   GetNArray(self, na);
   if (na->ndim == 0) {
@@ -4972,7 +5040,8 @@ static void iter_sfloat_min_index_index32_nan(na_loop_t* const lp) {
   Index of the minimum value.
   @overload min_index(axis:nil, nan:false)
   @param [TrueClass] nan  If true, apply NaN-aware algorithm (return NaN posision if exist).
-  @param [Numeric,Array,Range] axis  Finds minimum values along the axis and returns **flat 1-d indices**.
+  @param [Numeric,Array,Range] axis  Finds minimum values along the axis and returns **flat 1-d
+  indices**.
   @return [Integer,Numo::Int] returns result indices.
   @see #argmin
   @see #min
@@ -4990,9 +5059,9 @@ static void iter_sfloat_min_index_index32_nan(na_loop_t* const lp) {
 static VALUE sfloat_min_index(int argc, VALUE* argv, VALUE self) {
   narray_t* na;
   VALUE idx, reduce;
-  ndfunc_arg_in_t ain[3] = {{Qnil, 0}, {Qnil, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{0, 0, 0}};
-  ndfunc_t ndf = {0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { Qnil, 0 }, { Qnil, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { 0, 0, 0 } };
+  ndfunc_t ndf = { 0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 3, 1, ain, aout };
 
   GetNArray(self, na);
   if (na->ndim == 0) {
@@ -5085,7 +5154,8 @@ static void iter_sfloat_argmax_arg32_nan(na_loop_t* const lp) {
   Index of the maximum value.
   @overload argmax(axis:nil, nan:false)
   @param [TrueClass] nan  If true, apply NaN-aware algorithm (return NaN posision if exist).
-  @param [Numeric,Array,Range] axis  Finds maximum values along the axis and returns **indices along the axis**.
+  @param [Numeric,Array,Range] axis  Finds maximum values along the axis and returns **indices
+  along the axis**.
   @return [Integer,Numo::Int] returns the result indices.
   @see #max_index
   @see #max
@@ -5103,9 +5173,9 @@ static void iter_sfloat_argmax_arg32_nan(na_loop_t* const lp) {
 static VALUE sfloat_argmax(int argc, VALUE* argv, VALUE self) {
   narray_t* na;
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{Qnil, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{0, 0, 0}};
-  ndfunc_t ndf = {0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { Qnil, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { 0, 0, 0 } };
+  ndfunc_t ndf = { 0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 1, ain, aout };
 
   GetNArray(self, na);
   if (na->ndim == 0) {
@@ -5195,7 +5265,8 @@ static void iter_sfloat_argmin_arg32_nan(na_loop_t* const lp) {
   Index of the minimum value.
   @overload argmin(axis:nil, nan:false)
   @param [TrueClass] nan  If true, apply NaN-aware algorithm (return NaN posision if exist).
-  @param [Numeric,Array,Range] axis  Finds minimum values along the axis and returns **indices along the axis**.
+  @param [Numeric,Array,Range] axis  Finds minimum values along the axis and returns **indices
+  along the axis**.
   @return [Integer,Numo::Int] returns the result indices.
   @see #min_index
   @see #min
@@ -5213,9 +5284,9 @@ static void iter_sfloat_argmin_arg32_nan(na_loop_t* const lp) {
 static VALUE sfloat_argmin(int argc, VALUE* argv, VALUE self) {
   narray_t* na;
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{Qnil, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{0, 0, 0}};
-  ndfunc_t ndf = {0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { Qnil, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { 0, 0, 0 } };
+  ndfunc_t ndf = { 0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 1, ain, aout };
 
   GetNArray(self, na);
   if (na->ndim == 0) {
@@ -5271,14 +5342,17 @@ static void iter_sfloat_minmax_nan(na_loop_t* const lp) {
   @overload minmax(axis:nil, keepdims:false, nan:false)
   @param [TrueClass] nan  If true, apply NaN-aware algorithm (return NaN if exist).
   @param [Numeric,Array,Range] axis  Finds min-max along the axis.
-  @param [TrueClass] keepdims (keyword) If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims (keyword) If true, the reduced axes are left in the result array
+  as dimensions with size one.
   @return [Numo::SFloat,Numo::SFloat] min and max of self.
 */
 static VALUE sfloat_minmax(int argc, VALUE* argv, VALUE self) {
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_minmax, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 2, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_t ndf = {
+    iter_sfloat_minmax, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_EXTRACT, 2, 2, ain, aout
+  };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_minmax_nan);
 
@@ -5337,13 +5411,13 @@ static void iter_sfloat_s_maximum_nan(na_loop_t* const lp) {
 static VALUE sfloat_s_maximum(int argc, VALUE* argv, VALUE mod) {
   VALUE a1 = Qnil;
   VALUE a2 = Qnil;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_s_maximum, STRIDE_LOOP_NIP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_s_maximum, STRIDE_LOOP_NIP, 2, 1, ain, aout };
 
   VALUE kw_hash = Qnil;
-  ID kw_table[1] = {id_nan};
-  VALUE opts[1] = {Qundef};
+  ID kw_table[1] = { id_nan };
+  VALUE opts[1] = { Qundef };
 
   rb_scan_args(argc, argv, "20:", &a1, &a2, &kw_hash);
   rb_get_kwargs(kw_hash, kw_table, 0, 1, opts);
@@ -5406,13 +5480,13 @@ static void iter_sfloat_s_minimum_nan(na_loop_t* const lp) {
 static VALUE sfloat_s_minimum(int argc, VALUE* argv, VALUE mod) {
   VALUE a1 = Qnil;
   VALUE a2 = Qnil;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_s_minimum, STRIDE_LOOP_NIP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_s_minimum, STRIDE_LOOP_NIP, 2, 1, ain, aout };
 
   VALUE kw_hash = Qnil;
-  ID kw_table[1] = {id_nan};
-  VALUE opts[1] = {Qundef};
+  ID kw_table[1] = { id_nan };
+  VALUE opts[1] = { Qundef };
 
   rb_scan_args(argc, argv, "20:", &a1, &a2, &kw_hash);
   rb_get_kwargs(kw_hash, kw_table, 0, 1, opts);
@@ -5432,16 +5506,13 @@ static void iter_sfloat_cumsum(na_loop_t* const lp) {
   INIT_COUNTER(lp, i);
   INIT_PTR(lp, 0, p1, s1);
   INIT_PTR(lp, 1, p2, s2);
-  // printf("i=%lu p1=%lx s1=%lu p2=%lx s2=%lu\n",i,(size_t)p1,s1,(size_t)p2,s2);
 
   GET_DATA_STRIDE(p1, s1, dtype, x);
   SET_DATA_STRIDE(p2, s2, dtype, x);
-  // printf("i=%lu x=%f\n",i,x);
   for (i--; i--;) {
     GET_DATA_STRIDE(p1, s1, dtype, y);
     m_cumsum(x, y);
     SET_DATA_STRIDE(p2, s2, dtype, x);
-    // printf("i=%lu x=%f\n",i,x);
   }
 }
 static void iter_sfloat_cumsum_nan(na_loop_t* const lp) {
@@ -5453,16 +5524,13 @@ static void iter_sfloat_cumsum_nan(na_loop_t* const lp) {
   INIT_COUNTER(lp, i);
   INIT_PTR(lp, 0, p1, s1);
   INIT_PTR(lp, 1, p2, s2);
-  // printf("i=%lu p1=%lx s1=%lu p2=%lx s2=%lu\n",i,(size_t)p1,s1,(size_t)p2,s2);
 
   GET_DATA_STRIDE(p1, s1, dtype, x);
   SET_DATA_STRIDE(p2, s2, dtype, x);
-  // printf("i=%lu x=%f\n",i,x);
   for (i--; i--;) {
     GET_DATA_STRIDE(p1, s1, dtype, y);
     m_cumsum_nan(x, y);
     SET_DATA_STRIDE(p2, s2, dtype, x);
-    // printf("i=%lu x=%f\n",i,x);
   }
 }
 
@@ -5475,9 +5543,11 @@ static void iter_sfloat_cumsum_nan(na_loop_t* const lp) {
 */
 static VALUE sfloat_cumsum(int argc, VALUE* argv, VALUE self) {
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_cumsum, STRIDE_LOOP | NDF_FLAT_REDUCE | NDF_CUM, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = {
+    iter_sfloat_cumsum, STRIDE_LOOP | NDF_FLAT_REDUCE | NDF_CUM, 2, 1, ain, aout
+  };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_cumsum_nan);
 
@@ -5493,16 +5563,13 @@ static void iter_sfloat_cumprod(na_loop_t* const lp) {
   INIT_COUNTER(lp, i);
   INIT_PTR(lp, 0, p1, s1);
   INIT_PTR(lp, 1, p2, s2);
-  // printf("i=%lu p1=%lx s1=%lu p2=%lx s2=%lu\n",i,(size_t)p1,s1,(size_t)p2,s2);
 
   GET_DATA_STRIDE(p1, s1, dtype, x);
   SET_DATA_STRIDE(p2, s2, dtype, x);
-  // printf("i=%lu x=%f\n",i,x);
   for (i--; i--;) {
     GET_DATA_STRIDE(p1, s1, dtype, y);
     m_cumprod(x, y);
     SET_DATA_STRIDE(p2, s2, dtype, x);
-    // printf("i=%lu x=%f\n",i,x);
   }
 }
 static void iter_sfloat_cumprod_nan(na_loop_t* const lp) {
@@ -5514,16 +5581,13 @@ static void iter_sfloat_cumprod_nan(na_loop_t* const lp) {
   INIT_COUNTER(lp, i);
   INIT_PTR(lp, 0, p1, s1);
   INIT_PTR(lp, 1, p2, s2);
-  // printf("i=%lu p1=%lx s1=%lu p2=%lx s2=%lu\n",i,(size_t)p1,s1,(size_t)p2,s2);
 
   GET_DATA_STRIDE(p1, s1, dtype, x);
   SET_DATA_STRIDE(p2, s2, dtype, x);
-  // printf("i=%lu x=%f\n",i,x);
   for (i--; i--;) {
     GET_DATA_STRIDE(p1, s1, dtype, y);
     m_cumprod_nan(x, y);
     SET_DATA_STRIDE(p2, s2, dtype, x);
-    // printf("i=%lu x=%f\n",i,x);
   }
 }
 
@@ -5536,9 +5600,11 @@ static void iter_sfloat_cumprod_nan(na_loop_t* const lp) {
 */
 static VALUE sfloat_cumprod(int argc, VALUE* argv, VALUE self) {
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_cumprod, STRIDE_LOOP | NDF_FLAT_REDUCE | NDF_CUM, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = {
+    iter_sfloat_cumprod, STRIDE_LOOP | NDF_FLAT_REDUCE | NDF_CUM, 2, 1, ain, aout
+  };
 
   reduce = na_reduce_dimension(argc, argv, 1, &self, &ndf, iter_sfloat_cumprod_nan);
 
@@ -5618,9 +5684,9 @@ static void iter_sfloat_mulsum_nan(na_loop_t* const lp) {
 static VALUE sfloat_mulsum_self(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
   VALUE naryv[2];
-  ndfunc_arg_in_t ain[4] = {{cT, 0}, {cT, 0}, {sym_reduce, 0}, {sym_init, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_mulsum, STRIDE_LOOP_NIP, 4, 1, ain, aout};
+  ndfunc_arg_in_t ain[4] = { { cT, 0 }, { cT, 0 }, { sym_reduce, 0 }, { sym_init, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_mulsum, STRIDE_LOOP_NIP, 4, 1, ain, aout };
 
   if (argc < 1) {
     rb_raise(rb_eArgError, "wrong number of arguments (%d for >=1)", argc);
@@ -5642,7 +5708,8 @@ static VALUE sfloat_mulsum_self(int argc, VALUE* argv, VALUE self) {
   @overload mulsum(other, axis:nil, keepdims:false, nan:false)
   @param [Numo::NArray,Numeric] other
   @param [Numeric,Array,Range] axis  Performs mulsum along the axis.
-  @param [TrueClass] keepdims (keyword) If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims (keyword) If true, the reduced axes are left in the result array
+  as dimensions with size one.
   @param [TrueClass] nan (keyword) If true, apply NaN-aware algorithm (avoid NaN if exists).
   @return [Numo::NArray] mulsum of self and other.
 */
@@ -5728,8 +5795,8 @@ static void iter_sfloat_seq(na_loop_t* const lp) {
 static VALUE sfloat_seq(int argc, VALUE* args, VALUE self) {
   seq_opt_t* g;
   VALUE vbeg = Qnil, vstep = Qnil;
-  ndfunc_arg_in_t ain[1] = {{OVERWRITE, 0}};
-  ndfunc_t ndf = {iter_sfloat_seq, FULL_LOOP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { OVERWRITE, 0 } };
+  ndfunc_t ndf = { iter_sfloat_seq, FULL_LOOP, 1, 0, ain, 0 };
 
   g = ALLOCA_N(seq_opt_t, 1);
   g->beg = m_zero;
@@ -5811,8 +5878,8 @@ static void iter_sfloat_logseq(na_loop_t* const lp) {
 static VALUE sfloat_logseq(int argc, VALUE* args, VALUE self) {
   logseq_opt_t* g;
   VALUE vbeg, vstep, vbase;
-  ndfunc_arg_in_t ain[1] = {{OVERWRITE, 0}};
-  ndfunc_t ndf = {iter_sfloat_logseq, FULL_LOOP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { OVERWRITE, 0 } };
+  ndfunc_t ndf = { iter_sfloat_logseq, FULL_LOOP, 1, 0, ain, 0 };
 
   g = ALLOCA_N(logseq_opt_t, 1);
   rb_scan_args(argc, args, "21", &vbeg, &vstep, &vbase);
@@ -5866,8 +5933,8 @@ static void iter_sfloat_eye(na_loop_t* const lp) {
   @return [Numo::SFloat] eye of self.
 */
 static VALUE sfloat_eye(int argc, VALUE* argv, VALUE self) {
-  ndfunc_arg_in_t ain[1] = {{OVERWRITE, 2}};
-  ndfunc_t ndf = {iter_sfloat_eye, NO_LOOP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { OVERWRITE, 2 } };
+  ndfunc_t ndf = { iter_sfloat_eye, NO_LOOP, 1, 0, ain, 0 };
   ssize_t kofs;
   dtype data;
   char* g;
@@ -5897,17 +5964,21 @@ static VALUE sfloat_eye(int argc, VALUE* argv, VALUE self) {
   // Diagonal offset from the main diagonal.
   if (kofs >= 0) {
     if ((size_t)(kofs) >= na->shape[nd - 1]) {
-      rb_raise(rb_eArgError,
-               "invalid diagonal offset(%" SZF "d) for "
-               "last dimension size(%" SZF "d)",
-               kofs, na->shape[nd - 1]);
+      rb_raise(
+        rb_eArgError,
+        "invalid diagonal offset(%" SZF "d) for "
+        "last dimension size(%" SZF "d)",
+        kofs, na->shape[nd - 1]
+      );
     }
   } else {
     if ((size_t)(-kofs) >= na->shape[nd - 2]) {
-      rb_raise(rb_eArgError,
-               "invalid diagonal offset(%" SZF "d) for "
-               "last-1 dimension size(%" SZF "d)",
-               kofs, na->shape[nd - 2]);
+      rb_raise(
+        rb_eArgError,
+        "invalid diagonal offset(%" SZF "d) for "
+        "last-1 dimension size(%" SZF "d)",
+        kofs, na->shape[nd - 2]
+      );
     }
   }
 
@@ -5957,7 +6028,8 @@ static void iter_sfloat_rand(na_loop_t* const lp) {
   Generate uniformly distributed random numbers on self narray.
   @overload rand([[low],high])
   @param [Numeric] low  lower inclusive boundary of random numbers. (default=0)
-  @param [Numeric] high  upper exclusive boundary of random numbers. (default=1 or 1+1i for complex types)
+  @param [Numeric] high  upper exclusive boundary of random numbers. (default=1 or 1+1i for
+  complex types)
   @return [Numo::SFloat] self.
   @example
     Numo::DFloat.new(6).rand
@@ -5976,8 +6048,8 @@ static VALUE sfloat_rand(int argc, VALUE* args, VALUE self) {
   rand_opt_t g;
   VALUE v1 = Qnil, v2 = Qnil;
   dtype high;
-  ndfunc_arg_in_t ain[1] = {{OVERWRITE, 0}};
-  ndfunc_t ndf = {iter_sfloat_rand, FULL_LOOP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { OVERWRITE, 0 } };
+  ndfunc_t ndf = { iter_sfloat_rand, FULL_LOOP, 1, 0, ain, 0 };
 
   rb_scan_args(argc, args, "02", &v1, &v2);
   if (v2 == Qnil) {
@@ -6085,8 +6157,8 @@ static VALUE sfloat_rand_norm(int argc, VALUE* args, VALUE self) {
   int n;
   randn_opt_t g;
   VALUE v1 = Qnil, v2 = Qnil;
-  ndfunc_arg_in_t ain[1] = {{OVERWRITE, 0}};
-  ndfunc_t ndf = {iter_sfloat_rand_norm, FULL_LOOP, 1, 0, ain, 0};
+  ndfunc_arg_in_t ain[1] = { { OVERWRITE, 0 } };
+  ndfunc_t ndf = { iter_sfloat_rand_norm, FULL_LOOP, 1, 0, ain, 0 };
 
   n = rb_scan_args(argc, args, "02", &v1, &v2);
   if (n == 0) {
@@ -6130,8 +6202,8 @@ static VALUE sfloat_poly(VALUE self, VALUE args) {
   int argc, i;
   VALUE* argv;
   volatile VALUE v, a;
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_poly, NO_LOOP, 0, 1, 0, aout};
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_poly, NO_LOOP, 0, 1, 0, aout };
 
   argc = (int)RARRAY_LEN(args);
   ndf.nin = argc + 1;
@@ -6211,23 +6283,28 @@ static VALUE sfloat_poly(VALUE self, VALUE args) {
  * We have modified their original by adding a check for already-sorted input,
  * which seems to be a win per discussions on pgsql-hackers around 2006-03-21.
  */
-#define swapcode(TYPE, parmi, parmj, n)                                                                                        \
-  do {                                                                                                                         \
-    size_t i = (n) / sizeof(TYPE);                                                                                             \
-    TYPE* pi = (TYPE*)(void*)(parmi);                                                                                          \
-    TYPE* pj = (TYPE*)(void*)(parmj);                                                                                          \
-    do {                                                                                                                       \
-      TYPE t = *pi;                                                                                                            \
-      *pi++ = *pj;                                                                                                             \
-      *pj++ = t;                                                                                                               \
-    } while (--i > 0);                                                                                                         \
+#define swapcode(TYPE, parmi, parmj, n) \
+  do {                                  \
+    size_t i = (n) / sizeof(TYPE);      \
+    TYPE* pi = (TYPE*)(void*)(parmi);   \
+    TYPE* pj = (TYPE*)(void*)(parmj);   \
+    do {                                \
+      TYPE t = *pi;                     \
+      *pi++ = *pj;                      \
+      *pj++ = t;                        \
+    } while (--i > 0);                  \
   } while (0)
 
 #ifdef HAVE_STDINT_H
-#define SWAPINIT(a, es) swaptype = (uintptr_t)(a) % sizeof(long) || (es) % sizeof(long) ? 2 : (es) == sizeof(long) ? 0 : 1;
+#define SWAPINIT(a, es)                                               \
+  swaptype = (uintptr_t)(a) % sizeof(long) || (es) % sizeof(long) ? 2 \
+             : (es) == sizeof(long)                               ? 0 \
+                                                                  : 1;
 #else
-#define SWAPINIT(a, es)                                                                                                        \
-  swaptype = ((char*)(a) - (char*)0) % sizeof(long) || (es) % sizeof(long) ? 2 : (es) == sizeof(long) ? 0 : 1;
+#define SWAPINIT(a, es)                                                        \
+  swaptype = ((char*)(a) - (char*)0) % sizeof(long) || (es) % sizeof(long) ? 2 \
+             : (es) == sizeof(long)                                        ? 0 \
+                                                                           : 1;
 #endif
 
 static inline void swapfunc(char* a, char* b, size_t n, int swaptype) {
@@ -6237,19 +6314,20 @@ static inline void swapfunc(char* a, char* b, size_t n, int swaptype) {
     swapcode(char, a, b, n);
 }
 
-#define swap(a, b)                                                                                                             \
-  if (swaptype == 0) {                                                                                                         \
-    long t = *(long*)(void*)(a);                                                                                               \
-    *(long*)(void*)(a) = *(long*)(void*)(b);                                                                                   \
-    *(long*)(void*)(b) = t;                                                                                                    \
-  } else                                                                                                                       \
+#define swap(a, b)                           \
+  if (swaptype == 0) {                       \
+    long t = *(long*)(void*)(a);             \
+    *(long*)(void*)(a) = *(long*)(void*)(b); \
+    *(long*)(void*)(b) = t;                  \
+  } else                                     \
     swapfunc(a, b, es, swaptype)
 
-#define vecswap(a, b, n)                                                                                                       \
+#define vecswap(a, b, n) \
   if ((n) > 0) swapfunc((a), (b), (size_t)(n), swaptype)
 
-#define med3(a, b, c, _cmp)                                                                                                    \
-  (cmpgt(b, a) ? (cmpgt(c, b) ? b : (cmpgt(c, a) ? c : a)) : (cmpgt(b, c) ? b : (cmpgt(c, a) ? a : c)))
+#define med3(a, b, c, _cmp)                                \
+  (cmpgt(b, a) ? (cmpgt(c, b) ? b : (cmpgt(c, a) ? c : a)) \
+               : (cmpgt(b, c) ? b : (cmpgt(c, a) ? a : c)))
 #endif
 
 #undef qsort_dtype
@@ -6489,8 +6567,8 @@ static void iter_sfloat_sort_prnan(na_loop_t* const lp) {
 */
 static VALUE sfloat_sort(int argc, VALUE* argv, VALUE self) {
   VALUE reduce;
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {sym_reduce, 0}};
-  ndfunc_t ndf = {0, NDF_HAS_LOOP | NDF_FLAT_REDUCE, 2, 0, ain, 0};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { sym_reduce, 0 } };
+  ndfunc_t ndf = { 0, NDF_HAS_LOOP | NDF_FLAT_REDUCE, 2, 0, ain, 0 };
 
   if (!TEST_INPLACE(self)) {
     self = na_copy(self);
@@ -6595,7 +6673,8 @@ loop:
     pm = med3(pl, pm, pn, cmp);
   }
   swap(a, pm);
-  for (pa = pb = (char*)a + es, pc = pd = (char*)a + (n - 1) * es; pb <= pc; pb += es, pc -= es) {
+  for (pa = pb = (char*)a + es, pc = pd = (char*)a + (n - 1) * es; pb <= pc;
+       pb += es, pc -= es) {
     while (pb <= pc && (r = cmp(pb, a)) <= 0) {
       if (r == 0) {
         swap(pa, pb);
@@ -6720,7 +6799,8 @@ loop:
     pm = med3(pl, pm, pn, cmp);
   }
   swap(a, pm);
-  for (pa = pb = (char*)a + es, pc = pd = (char*)a + (n - 1) * es; pb <= pc; pb += es, pc -= es) {
+  for (pa = pb = (char*)a + es, pc = pd = (char*)a + (n - 1) * es; pb <= pc;
+       pb += es, pc -= es) {
     while (pb <= pc && (r = cmp(pb, a)) <= 0) {
       if (r == 0) {
         swap(pa, pb);
@@ -6767,7 +6847,6 @@ static void sfloat_index64_qsort_ignan(na_loop_t* const lp) {
 
   ptr = (char**)(lp->opt_ptr);
 
-  // printf("(ptr=%lx, d_ptr=%lx,d_step=%ld, i_ptr=%lx,i_step=%ld,
   // o_ptr=%lx,o_step=%ld)\n",(size_t)ptr,(size_t)d_ptr,(ssize_t)d_step,(size_t)i_ptr,(ssize_t)i_step,(size_t)o_ptr,(ssize_t)o_step);
 
   if (n == 1) {
@@ -6777,21 +6856,17 @@ static void sfloat_index64_qsort_ignan(na_loop_t* const lp) {
 
   for (i = 0; i < n; i++) {
     ptr[i] = d_ptr + d_step * i;
-    // printf("(%ld,%.3f)",i,*(double*)ptr[i]);
   }
 
   sfloat_index_qsort_ignan(ptr, n, sizeof(dtype*));
 
   // d_ptr = lp->args[0].ptr;
-  // printf("(d_ptr=%lx)\n",(size_t)d_ptr);
 
   for (i = 0; i < n; i++) {
     idx = (ptr[i] - d_ptr) / d_step;
     *(idx_t*)o_ptr = *(idx_t*)(i_ptr + i_step * idx);
-    // printf("(idx[%ld]=%ld,%d)",i,idx,*(idx_t*)o_ptr);
     o_ptr += o_step;
   }
-  // printf("\n");
 }
 #undef idx_t
 
@@ -6809,7 +6884,6 @@ static void sfloat_index32_qsort_ignan(na_loop_t* const lp) {
 
   ptr = (char**)(lp->opt_ptr);
 
-  // printf("(ptr=%lx, d_ptr=%lx,d_step=%ld, i_ptr=%lx,i_step=%ld,
   // o_ptr=%lx,o_step=%ld)\n",(size_t)ptr,(size_t)d_ptr,(ssize_t)d_step,(size_t)i_ptr,(ssize_t)i_step,(size_t)o_ptr,(ssize_t)o_step);
 
   if (n == 1) {
@@ -6819,21 +6893,17 @@ static void sfloat_index32_qsort_ignan(na_loop_t* const lp) {
 
   for (i = 0; i < n; i++) {
     ptr[i] = d_ptr + d_step * i;
-    // printf("(%ld,%.3f)",i,*(double*)ptr[i]);
   }
 
   sfloat_index_qsort_ignan(ptr, n, sizeof(dtype*));
 
   // d_ptr = lp->args[0].ptr;
-  // printf("(d_ptr=%lx)\n",(size_t)d_ptr);
 
   for (i = 0; i < n; i++) {
     idx = (ptr[i] - d_ptr) / d_step;
     *(idx_t*)o_ptr = *(idx_t*)(i_ptr + i_step * idx);
-    // printf("(idx[%ld]=%ld,%d)",i,idx,*(idx_t*)o_ptr);
     o_ptr += o_step;
   }
-  // printf("\n");
 }
 #undef idx_t
 
@@ -6851,7 +6921,6 @@ static void sfloat_index64_qsort_prnan(na_loop_t* const lp) {
 
   ptr = (char**)(lp->opt_ptr);
 
-  // printf("(ptr=%lx, d_ptr=%lx,d_step=%ld, i_ptr=%lx,i_step=%ld,
   // o_ptr=%lx,o_step=%ld)\n",(size_t)ptr,(size_t)d_ptr,(ssize_t)d_step,(size_t)i_ptr,(ssize_t)i_step,(size_t)o_ptr,(ssize_t)o_step);
 
   if (n == 1) {
@@ -6861,21 +6930,17 @@ static void sfloat_index64_qsort_prnan(na_loop_t* const lp) {
 
   for (i = 0; i < n; i++) {
     ptr[i] = d_ptr + d_step * i;
-    // printf("(%ld,%.3f)",i,*(double*)ptr[i]);
   }
 
   sfloat_index_qsort_prnan(ptr, n, sizeof(dtype*));
 
   // d_ptr = lp->args[0].ptr;
-  // printf("(d_ptr=%lx)\n",(size_t)d_ptr);
 
   for (i = 0; i < n; i++) {
     idx = (ptr[i] - d_ptr) / d_step;
     *(idx_t*)o_ptr = *(idx_t*)(i_ptr + i_step * idx);
-    // printf("(idx[%ld]=%ld,%d)",i,idx,*(idx_t*)o_ptr);
     o_ptr += o_step;
   }
-  // printf("\n");
 }
 #undef idx_t
 
@@ -6893,7 +6958,6 @@ static void sfloat_index32_qsort_prnan(na_loop_t* const lp) {
 
   ptr = (char**)(lp->opt_ptr);
 
-  // printf("(ptr=%lx, d_ptr=%lx,d_step=%ld, i_ptr=%lx,i_step=%ld,
   // o_ptr=%lx,o_step=%ld)\n",(size_t)ptr,(size_t)d_ptr,(ssize_t)d_step,(size_t)i_ptr,(ssize_t)i_step,(size_t)o_ptr,(ssize_t)o_step);
 
   if (n == 1) {
@@ -6903,21 +6967,17 @@ static void sfloat_index32_qsort_prnan(na_loop_t* const lp) {
 
   for (i = 0; i < n; i++) {
     ptr[i] = d_ptr + d_step * i;
-    // printf("(%ld,%.3f)",i,*(double*)ptr[i]);
   }
 
   sfloat_index_qsort_prnan(ptr, n, sizeof(dtype*));
 
   // d_ptr = lp->args[0].ptr;
-  // printf("(d_ptr=%lx)\n",(size_t)d_ptr);
 
   for (i = 0; i < n; i++) {
     idx = (ptr[i] - d_ptr) / d_step;
     *(idx_t*)o_ptr = *(idx_t*)(i_ptr + i_step * idx);
-    // printf("(idx[%ld]=%ld,%d)",i,idx,*(idx_t*)o_ptr);
     o_ptr += o_step;
   }
-  // printf("\n");
 }
 #undef idx_t
 
@@ -6935,9 +6995,9 @@ static VALUE sfloat_sort_index(int argc, VALUE* argv, VALUE self) {
   narray_t* na;
   VALUE idx, tmp, reduce, res;
   char* buf;
-  ndfunc_arg_in_t ain[3] = {{cT, 0}, {0, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{0, 0, 0}};
-  ndfunc_t ndf = {0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_CUM, 3, 1, ain, aout};
+  ndfunc_arg_in_t ain[3] = { { cT, 0 }, { 0, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { 0, 0, 0 } };
+  ndfunc_t ndf = { 0, STRIDE_LOOP_NIP | NDF_FLAT_REDUCE | NDF_CUM, 3, 1, ain, aout };
 
   GetNArray(self, na);
   if (na->ndim == 0) {
@@ -7020,15 +7080,16 @@ static void iter_sfloat_median_prnan(na_loop_t* const lp) {
   @overload median(axis:nil, keepdims:false, nan:false)
   @param [TrueClass] nan (keyword) If true, propagete NaN. If false, ignore NaN.
   @param [Numeric,Array,Range] axis  Finds median along the axis.
-  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as dimensions with size one.
+  @param [TrueClass] keepdims  If true, the reduced axes are left in the result array as
+  dimensions with size one.
   @return [Numo::SFloat] returns median of self.
 */
 
 static VALUE sfloat_median(int argc, VALUE* argv, VALUE self) {
   VALUE v, reduce;
-  ndfunc_arg_in_t ain[2] = {{OVERWRITE, 0}, {sym_reduce, 0}};
-  ndfunc_arg_out_t aout[1] = {{INT2FIX(0), 0}};
-  ndfunc_t ndf = {0, NDF_HAS_LOOP | NDF_FLAT_REDUCE, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { OVERWRITE, 0 }, { sym_reduce, 0 } };
+  ndfunc_arg_out_t aout[1] = { { INT2FIX(0), 0 } };
+  ndfunc_t ndf = { 0, NDF_HAS_LOOP | NDF_FLAT_REDUCE, 2, 1, ain, aout };
 
   self = na_copy(self); // as temporary buffer
 
@@ -7093,9 +7154,12 @@ static void iter_sfloat_math_s_sqrt(na_loop_t* const lp) {
 #ifdef __SSE2__
           //
           // Check number of elements. & Check same alignment.
-          if ((n >= num_pack) && is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p2)[i], SIMD_ALIGNMENT_SIZE)) {
+          if ((n >= num_pack) &&
+              is_same_aligned2(&((dtype*)p1)[i], &((dtype*)p2)[i], SIMD_ALIGNMENT_SIZE)) {
             // Calculate up to the position just before the start of SIMD computation.
-            cnt = get_count_of_elements_not_aligned_to_simd_size(&((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype));
+            cnt = get_count_of_elements_not_aligned_to_simd_size(
+              &((dtype*)p1)[i], SIMD_ALIGNMENT_SIZE, sizeof(dtype)
+            );
 #endif
 #ifdef __SSE2__
             for (i = 0; i < cnt; i++) {
@@ -7164,9 +7228,9 @@ static void iter_sfloat_math_s_sqrt(na_loop_t* const lp) {
   @return [Numo::SFloat] result of sqrt(x).
 */
 static VALUE sfloat_math_s_sqrt(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_sqrt, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_sqrt, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7242,9 +7306,9 @@ static void iter_sfloat_math_s_cbrt(na_loop_t* const lp) {
   @return [Numo::SFloat] result of cbrt(x).
 */
 static VALUE sfloat_math_s_cbrt(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_cbrt, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_cbrt, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7320,9 +7384,9 @@ static void iter_sfloat_math_s_log(na_loop_t* const lp) {
   @return [Numo::SFloat] result of log(x).
 */
 static VALUE sfloat_math_s_log(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_log, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_log, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7398,9 +7462,9 @@ static void iter_sfloat_math_s_log2(na_loop_t* const lp) {
   @return [Numo::SFloat] result of log2(x).
 */
 static VALUE sfloat_math_s_log2(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_log2, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_log2, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7476,9 +7540,9 @@ static void iter_sfloat_math_s_log10(na_loop_t* const lp) {
   @return [Numo::SFloat] result of log10(x).
 */
 static VALUE sfloat_math_s_log10(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_log10, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_log10, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7554,9 +7618,9 @@ static void iter_sfloat_math_s_exp(na_loop_t* const lp) {
   @return [Numo::SFloat] result of exp(x).
 */
 static VALUE sfloat_math_s_exp(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_exp, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_exp, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7632,9 +7696,9 @@ static void iter_sfloat_math_s_exp2(na_loop_t* const lp) {
   @return [Numo::SFloat] result of exp2(x).
 */
 static VALUE sfloat_math_s_exp2(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_exp2, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_exp2, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7710,9 +7774,9 @@ static void iter_sfloat_math_s_exp10(na_loop_t* const lp) {
   @return [Numo::SFloat] result of exp10(x).
 */
 static VALUE sfloat_math_s_exp10(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_exp10, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_exp10, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7788,9 +7852,9 @@ static void iter_sfloat_math_s_sin(na_loop_t* const lp) {
   @return [Numo::SFloat] result of sin(x).
 */
 static VALUE sfloat_math_s_sin(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_sin, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_sin, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7866,9 +7930,9 @@ static void iter_sfloat_math_s_cos(na_loop_t* const lp) {
   @return [Numo::SFloat] result of cos(x).
 */
 static VALUE sfloat_math_s_cos(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_cos, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_cos, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -7944,9 +8008,9 @@ static void iter_sfloat_math_s_tan(na_loop_t* const lp) {
   @return [Numo::SFloat] result of tan(x).
 */
 static VALUE sfloat_math_s_tan(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_tan, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_tan, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8022,9 +8086,9 @@ static void iter_sfloat_math_s_asin(na_loop_t* const lp) {
   @return [Numo::SFloat] result of asin(x).
 */
 static VALUE sfloat_math_s_asin(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_asin, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_asin, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8100,9 +8164,9 @@ static void iter_sfloat_math_s_acos(na_loop_t* const lp) {
   @return [Numo::SFloat] result of acos(x).
 */
 static VALUE sfloat_math_s_acos(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_acos, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_acos, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8178,9 +8242,9 @@ static void iter_sfloat_math_s_atan(na_loop_t* const lp) {
   @return [Numo::SFloat] result of atan(x).
 */
 static VALUE sfloat_math_s_atan(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_atan, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_atan, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8256,9 +8320,9 @@ static void iter_sfloat_math_s_sinh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of sinh(x).
 */
 static VALUE sfloat_math_s_sinh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_sinh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_sinh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8334,9 +8398,9 @@ static void iter_sfloat_math_s_cosh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of cosh(x).
 */
 static VALUE sfloat_math_s_cosh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_cosh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_cosh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8412,9 +8476,9 @@ static void iter_sfloat_math_s_tanh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of tanh(x).
 */
 static VALUE sfloat_math_s_tanh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_tanh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_tanh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8490,9 +8554,9 @@ static void iter_sfloat_math_s_asinh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of asinh(x).
 */
 static VALUE sfloat_math_s_asinh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_asinh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_asinh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8568,9 +8632,9 @@ static void iter_sfloat_math_s_acosh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of acosh(x).
 */
 static VALUE sfloat_math_s_acosh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_acosh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_acosh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8646,9 +8710,9 @@ static void iter_sfloat_math_s_atanh(na_loop_t* const lp) {
   @return [Numo::SFloat] result of atanh(x).
 */
 static VALUE sfloat_math_s_atanh(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_atanh, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_atanh, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8724,9 +8788,9 @@ static void iter_sfloat_math_s_sinc(na_loop_t* const lp) {
   @return [Numo::SFloat] result of sinc(x).
 */
 static VALUE sfloat_math_s_sinc(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_sinc, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_sinc, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8756,9 +8820,9 @@ static void iter_sfloat_math_s_atan2(na_loop_t* const lp) {
   @return [Numo::SFloat] atan2(a1,a2).
 */
 static VALUE sfloat_math_s_atan2(VALUE mod, VALUE a1, VALUE a2) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_atan2, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_atan2, STRIDE_LOOP, 2, 1, ain, aout };
   return na_ndloop(&ndf, 2, a1, a2);
 }
 
@@ -8787,9 +8851,9 @@ static void iter_sfloat_math_s_hypot(na_loop_t* const lp) {
   @return [Numo::SFloat] hypot(a1,a2).
 */
 static VALUE sfloat_math_s_hypot(VALUE mod, VALUE a1, VALUE a2) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_hypot, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_hypot, STRIDE_LOOP, 2, 1, ain, aout };
   return na_ndloop(&ndf, 2, a1, a2);
 }
 
@@ -8864,9 +8928,9 @@ static void iter_sfloat_math_s_erf(na_loop_t* const lp) {
   @return [Numo::SFloat] result of erf(x).
 */
 static VALUE sfloat_math_s_erf(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_erf, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_erf, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -8942,9 +9006,9 @@ static void iter_sfloat_math_s_erfc(na_loop_t* const lp) {
   @return [Numo::SFloat] result of erfc(x).
 */
 static VALUE sfloat_math_s_erfc(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_erfc, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_erfc, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -9020,9 +9084,9 @@ static void iter_sfloat_math_s_log1p(na_loop_t* const lp) {
   @return [Numo::SFloat] result of log1p(x).
 */
 static VALUE sfloat_math_s_log1p(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_log1p, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_log1p, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -9098,9 +9162,9 @@ static void iter_sfloat_math_s_expm1(na_loop_t* const lp) {
   @return [Numo::SFloat] result of expm1(x).
 */
 static VALUE sfloat_math_s_expm1(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_expm1, FULL_LOOP, 1, 1, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_expm1, FULL_LOOP, 1, 1, ain, aout };
 
   return na_ndloop(&ndf, 1, a1);
 }
@@ -9130,9 +9194,9 @@ static void iter_sfloat_math_s_ldexp(na_loop_t* const lp) {
   @return [Numo::SFloat] ldexp(a1,a2).
 */
 static VALUE sfloat_math_s_ldexp(VALUE mod, VALUE a1, VALUE a2) {
-  ndfunc_arg_in_t ain[2] = {{cT, 0}, {cT, 0}};
-  ndfunc_arg_out_t aout[1] = {{cT, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_ldexp, STRIDE_LOOP, 2, 1, ain, aout};
+  ndfunc_arg_in_t ain[2] = { { cT, 0 }, { cT, 0 } };
+  ndfunc_arg_out_t aout[1] = { { cT, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_ldexp, STRIDE_LOOP, 2, 1, ain, aout };
   return na_ndloop(&ndf, 2, a1, a2);
 }
 
@@ -9164,9 +9228,9 @@ static void iter_sfloat_math_s_frexp(na_loop_t* const lp) {
 
 */
 static VALUE sfloat_math_s_frexp(VALUE mod, VALUE a1) {
-  ndfunc_arg_in_t ain[1] = {{cT, 0}};
-  ndfunc_arg_out_t aout[2] = {{cT, 0}, {numo_cInt32, 0}};
-  ndfunc_t ndf = {iter_sfloat_math_s_frexp, STRIDE_LOOP, 1, 2, ain, aout};
+  ndfunc_arg_in_t ain[1] = { { cT, 0 } };
+  ndfunc_arg_out_t aout[2] = { { cT, 0 }, { numo_cInt32, 0 } };
+  ndfunc_t ndf = { iter_sfloat_math_s_frexp, STRIDE_LOOP, 1, 2, ain, aout };
   return na_ndloop(&ndf, 1, a1);
 }