From 8cd5a3a3a99cc6a6ee6af228a717c1c3f2542c73 Mon Sep 17 00:00:00 2001
From: server2 <server2@server2.server2>
Date: Sun, 8 Dec 2024 16:37:51 -0500
Subject: [PATCH] Sun Dec  8 04:37:51 PM EST 2024

---
 src/bytecode.c.measurement.backup | 1751 +++++++++++++++++++++++++++++
 1 file changed, 1751 insertions(+)
 create mode 100644 src/bytecode.c.measurement.backup

diff --git a/src/bytecode.c.measurement.backup b/src/bytecode.c.measurement.backup
new file mode 100644
index 0000000..8212add
--- /dev/null
+++ b/src/bytecode.c.measurement.backup
@@ -0,0 +1,1751 @@
+#include "bytecode.h"
+
+const char* qansel_instruction_to_string(unsigned char instr)
+{
+	switch (instr)
+	{
+		case QANSEL_INSTRUCTION_X: return "x";
+		case QANSEL_INSTRUCTION_Y: return "y";
+		case QANSEL_INSTRUCTION_Z: return "z";
+		case QANSEL_INSTRUCTION_H: return "h";
+		case QANSEL_INSTRUCTION_S: return "s";
+		case QANSEL_INSTRUCTION_T: return "t";
+		case QANSEL_INSTRUCTION_SDG: return "sdg";
+		case QANSEL_INSTRUCTION_TDG: return "tdg";
+		case QANSEL_INSTRUCTION_RX: return "rx";
+		case QANSEL_INSTRUCTION_RY: return "ry";
+		case QANSEL_INSTRUCTION_RZ: return "rz";
+		case QANSEL_INSTRUCTION_U1: return "u1";
+		case QANSEL_INSTRUCTION_U2: return "u2";
+		case QANSEL_INSTRUCTION_U3: return "u3";
+		case QANSEL_INSTRUCTION_CX: return "cx";
+		case QANSEL_INSTRUCTION_SWAP: return "swap";
+		case QANSEL_INSTRUCTION_CCX: return "ccx";
+		case QANSEL_INSTRUCTION_CSWAP: return "cswap";
+		case QANSEL_INSTRUCTION_MEASURE: return "measure";
+		case QANSEL_INSTRUCTION_DENSITY: return "density";
+		case QANSEL_INSTRUCTION_BORN: return "born";
+		case QANSEL_INSTRUCTION_BSAMPLE: return "bsample";
+		case QANSEL_INSTRUCTION_IF_E: return "if==";
+		case QANSEL_INSTRUCTION_IF_NE: return "if!=";
+		case QANSEL_INSTRUCTION_IF_G: return "if>";
+		case QANSEL_INSTRUCTION_IF_GE: return "if>=";
+		case QANSEL_INSTRUCTION_IF_L: return "if<";
+		case QANSEL_INSTRUCTION_IF_LE: return "if<=";
+		case QANSEL_INSTRUCTION_PRINT: return "print";
+		case QANSEL_INSTRUCTION_SET: return "set";
+		case QANSEL_INSTRUCTION_RESET: return "reset";
+		case QANSEL_INSTRUCTION_BARRIER: return "barrier";
+		case QANSEL_INSTRUCTION_EXIT: return "exit";
+	}
+	return "Unknown";
+}
+
+float qansel_rand_s(float s)
+{
+	unsigned int tmp;
+	memcpy(&tmp, &s, sizeof(unsigned int));
+	srand(tmp);
+}
+float qansel_rand_h()
+{
+	return ((float)rand()) / ((float)RAND_MAX);
+}
+float qansel_rand_t(QAnselContext* ctx)
+{
+	if (ctx->hardware_rng != QANSEL_HARDWARE_NONE)
+	{
+		unsigned int num = 0;
+		for (unsigned char i = 0; i < 4; i++)
+		{
+			unsigned char r = 0;
+			switch (ctx->hardware_rng)
+			{
+				case QANSEL_HARDWARE_TRUERNG: r = fgetc(ctx->random_file); break;
+				case QANSEL_HARDWARE_QUANTIS:
+					if (ctx->pointer >= QANSEL_QUANTIS_CHUNK_SIZE)
+					{
+						qansel_quantis_chunk(&(ctx->chunk));
+						ctx->pointer = 0;
+					}
+					r = ctx->chunk[ctx->pointer];
+					ctx->pointer += 1;
+				break;
+				case QANSEL_HARDWARE_RDSEED: r = qansel_hardware_rand(); break;
+			}
+			num = (num << 8) | r;
+		}
+		return ((float)num) / ((float)UINT32_MAX);
+	}
+	else
+	{
+		return qansel_rand_h();
+	}
+}
+
+void qansel_cnot(cpx_mtx_t* stateVector, unsigned char qubitCount, unsigned char bitA, unsigned char bitB)
+{
+	if (bitA >= qubitCount || bitB >= qubitCount) return;
+	unsigned int retLen = (unsigned int)pow(2, qubitCount);
+	cpx_mtx_t ret;
+	cpx_mtx_init(&ret, 1, retLen);
+	cpx_t n;
+	for (unsigned int i = 0; i < retLen; i++)
+	{
+		unsigned char bitAVal = (i >> bitA) & 1;
+		unsigned char bitBVal = (i >> bitB) & 1;
+		unsigned char bitBNew = bitAVal ? !bitBVal : bitBVal;
+		unsigned int j = (i & ~(1 << bitB)) | (bitBNew << bitB);
+		cpx_mtx_get(stateVector, 0, i, &n);
+		cpx_mtx_set(&ret, 0, j, &n);
+	}
+	cpx_mtx_free(stateVector);
+	stateVector->ptr = ret.ptr;
+	stateVector->rows = ret.rows;
+	stateVector->cols = ret.cols;
+}
+
+void qansel_swap(cpx_mtx_t* stateVector, unsigned char qubitCount, unsigned char bitA, unsigned char bitB)
+{
+	if (bitA >= qubitCount || bitB >= qubitCount) return;
+	unsigned int retLen = (unsigned int)pow(2, qubitCount);
+	cpx_mtx_t ret;
+	cpx_mtx_init(&ret, 1, retLen);
+	cpx_t n;
+	for (unsigned int i = 0; i < retLen; i++)
+	{
+		unsigned char bitAVal = (i >> bitA) & 1;
+		unsigned char bitBVal = (i >> bitB) & 1;
+		unsigned char bitANew = bitBVal;
+		unsigned char bitBNew = bitAVal;
+		unsigned int j = (i & ~((1 << bitA) | (1 << bitB))) | ((bitANew << bitA) | (bitBNew << bitB));
+		cpx_mtx_get(stateVector, 0, i, &n);
+		cpx_mtx_set(&ret, 0, j, &n);
+	}
+	cpx_mtx_free(stateVector);
+	stateVector->ptr = ret.ptr;
+	stateVector->rows = ret.rows;
+	stateVector->cols = ret.cols;
+}
+
+void qansel_fredkin(cpx_mtx_t* stateVector, unsigned char qubitCount, unsigned char bitA, unsigned char bitB, unsigned char bitC)
+{
+	if (bitA >= qubitCount || bitB >= qubitCount) return;
+	unsigned int retLen = (unsigned int)pow(2, qubitCount);
+	cpx_mtx_t ret;
+	cpx_mtx_init(&ret, 1, retLen);
+	cpx_t n;
+	for (unsigned int i = 0; i < retLen; i++)
+	{
+		unsigned char bitAVal = (i >> bitA) & 1;
+		unsigned char bitBVal = (i >> bitB) & 1;
+		unsigned char bitCVal = (i >> bitC) & 1;
+		unsigned char bitBNew = bitAVal ? bitCVal : bitBVal;
+		unsigned char bitCNew = bitAVal ? bitBVal : bitCVal;
+		unsigned int j = (i & ~((1 << bitB) | (1 << bitC))) | ((bitBNew << bitB) | (bitCNew << bitC));
+		cpx_mtx_get(stateVector, 0, i, &n);
+		cpx_mtx_set(&ret, 0, j, &n);
+	}
+	cpx_mtx_free(stateVector);
+	stateVector->ptr = ret.ptr;
+	stateVector->rows = ret.rows;
+	stateVector->cols = ret.cols;
+}
+
+void qansel_toffoli(cpx_mtx_t* stateVector, unsigned char qubitCount, unsigned char bitA, unsigned char bitB, unsigned char bitC)
+{
+	if (bitA >= qubitCount || bitB >= qubitCount) return;
+	unsigned int retLen = (unsigned int)pow(2, qubitCount);
+	cpx_mtx_t ret;
+	cpx_mtx_init(&ret, 1, retLen);
+	cpx_t n;
+	for (unsigned int i = 0; i < retLen; i++)
+	{
+		unsigned char bitAVal = (i >> bitA) & 1;
+		unsigned char bitBVal = (i >> bitB) & 1;
+		unsigned char bitCVal = (i >> bitC) & 1;
+		unsigned char bitCNew = (bitAVal && bitBVal) ? !bitCVal : bitCVal;
+		unsigned int j = (i & ~(1 << bitC)) | (bitCNew << bitC);
+		cpx_mtx_get(stateVector, 0, i, &n);
+		cpx_mtx_set(&ret, 0, j, &n);
+	}
+	cpx_mtx_free(stateVector);
+	stateVector->ptr = ret.ptr;
+	stateVector->rows = ret.rows;
+	stateVector->cols = ret.cols;
+}
+
+float* qansel_unitary(float theta, float phi, float lambda)
+{
+	cpx_mtx_t m;
+	cpx_t a, b, c, d;
+	a.real = cos(theta/2.0);
+	a.imaginary = 0;
+	b.real = -cos(lambda) * sin(theta/2.0);
+	b.imaginary = sin(lambda) * sin(theta/2.0);
+	c.real = cos(phi) * sin(theta/2.0);
+	c.imaginary = sin(phi) * sin(theta/2.0);
+	d.real = cos(phi + lambda) * cos(theta/2.0);
+	d.imaginary = sin(phi + lambda) * cos(theta/2.0);
+	//remember it must be transposed!!!
+	cpx_mtx_init(&m, 2, 2);
+	cpx_mtx_set(&m, 0, 0, &a);
+	cpx_mtx_set(&m, 0, 1, &c);
+	cpx_mtx_set(&m, 1, 0, &b);
+	cpx_mtx_set(&m, 1, 1, &d);
+	return m.ptr;
+}
+
+void qansel_queue_init
+(
+	cpx_mtx_t* queueVector,
+	int qubitCount,
+	int dofree
+)
+{
+	for (int i = 0; i < qubitCount; i++)
+	{
+		if (dofree) free(queueVector[i].ptr);
+		int vectLen = queueVector[i].rows * (queueVector[i].cols * 2) + queueVector[i].cols * 2;
+		queueVector[i].rows = 2;
+		queueVector[i].cols = 2;
+		queueVector[i].ptr = malloc(vectLen);
+		memcpy(queueVector[i].ptr, Identity, vectLen);
+	}
+}
+
+void qansel_instruction_queue
+(
+	cpx_mtx_t** queueVector,
+	int qubitCount,
+	unsigned char instr,
+	unsigned char index, 
+	float arg0,
+	float arg1,
+	float arg2
+)
+{
+	cpx_mtx_t gate, tmp;
+	gate.rows = 2;
+	gate.cols = 2;
+	float* gate_ptr;
+	int needToFreeGate = 0;
+	switch (instr)
+	{
+		case QANSEL_INSTRUCTION_H: gate_ptr = Hadamard; break;
+		case QANSEL_INSTRUCTION_X: gate_ptr = PauliX; break;
+		case QANSEL_INSTRUCTION_Y: gate_ptr = PauliY; break;
+		case QANSEL_INSTRUCTION_Z: gate_ptr = PauliZ; break;
+		case QANSEL_INSTRUCTION_S: gate_ptr = PhaseS; break;
+		case QANSEL_INSTRUCTION_T: gate_ptr = PhaseT; break;
+		case QANSEL_INSTRUCTION_SDG: gate_ptr = PhaseSdg; break;
+		case QANSEL_INSTRUCTION_TDG: gate_ptr = PhaseTdg; break;
+		case QANSEL_INSTRUCTION_RX:
+		case QANSEL_INSTRUCTION_RY:
+		case QANSEL_INSTRUCTION_RZ:
+		case QANSEL_INSTRUCTION_U1:
+		case QANSEL_INSTRUCTION_U2:
+		case QANSEL_INSTRUCTION_U3:
+			gate_ptr = qansel_unitary(arg0, arg1, arg2);
+			needToFreeGate = 1;
+			break;
+		default: gate_ptr = Identity; break;
+	}
+	unsigned char qubit = qubitCount - (index) - 1;
+	gate.rows = 2;
+	gate.cols = 2;
+	gate.ptr = gate_ptr;
+	tmp.rows = 2;
+	tmp.cols = 2;
+	tmp.ptr = malloc((tmp.rows * 2) * (tmp.cols * 2) * sizeof(float));
+	cpx_mtx_dot(tmp.ptr, (*queueVector)[qubit].ptr, gate_ptr, (*queueVector)[qubit].rows, (*queueVector)[qubit].cols, 2, 2);
+	free((*queueVector)[qubit].ptr);
+	(*queueVector)[qubit].rows = tmp.rows;
+	(*queueVector)[qubit].cols = tmp.cols;
+	(*queueVector)[qubit].ptr = tmp.ptr;
+	if (needToFreeGate) free(gate_ptr);
+}
+
+void qansel_instruction
+(
+	QAnselContext* ctx,
+	cpx_mtx_t* stateVector,
+	int qubitCount,
+	unsigned char instr,
+	unsigned char index, 
+	float arg0,
+	float arg1,
+	float arg2,
+	//pass in this to flush the queue
+	cpx_mtx_t** queueVector
+)
+{
+	cpx_mtx_t tmp;
+	cpx_mtx_t gate;
+	gate.rows = 2;
+	gate.cols = 2;
+	float* gate_ptr;
+	int needToFreeGate = 0;
+	switch (instr)
+	{
+		case QANSEL_INSTRUCTION_H: gate_ptr = Hadamard; break;
+		case QANSEL_INSTRUCTION_X: gate_ptr = PauliX; break;
+		case QANSEL_INSTRUCTION_Y: gate_ptr = PauliY; break;
+		case QANSEL_INSTRUCTION_Z: gate_ptr = PauliZ; break;
+		case QANSEL_INSTRUCTION_S: gate_ptr = PhaseS; break;
+		case QANSEL_INSTRUCTION_T: gate_ptr = PhaseT; break;
+		case QANSEL_INSTRUCTION_SDG: gate_ptr = PhaseSdg; break;
+		case QANSEL_INSTRUCTION_TDG: gate_ptr = PhaseTdg; break;
+		case QANSEL_INSTRUCTION_RX:
+		case QANSEL_INSTRUCTION_RY:
+		case QANSEL_INSTRUCTION_RZ:
+		case QANSEL_INSTRUCTION_U1:
+		case QANSEL_INSTRUCTION_U2:
+		case QANSEL_INSTRUCTION_U3:
+			gate_ptr = qansel_unitary(arg0, arg1, arg2);
+			needToFreeGate = 1;
+			break;
+		default: gate_ptr = Identity; break;
+	}
+
+	cpx_t n;
+	cpx_mtx_t filter;
+	cpx_mtx_init(&filter, 2, 2);
+	unsigned char qubit = qubitCount - (queueVector == NULL ? index : 0) - 1;
+	if (qubit == 0 && queueVector == NULL)
+	{
+		memcpy(filter.ptr, gate_ptr, 8 * sizeof(float));
+	}
+	else if (queueVector == NULL)
+	{
+		memcpy(filter.ptr, Identity, 8 * sizeof(float));
+	}
+	else
+	{
+		memcpy(filter.ptr, (*queueVector)[0].ptr, 8 * sizeof(float));
+	}
+
+	for (unsigned char i = 1; i < qubitCount; i++)
+	{
+		if (index != QANSEL_ALL_QUANTUM && queueVector == NULL)
+		{
+			if (qubit == i)
+			{
+				gate.ptr = gate_ptr;
+			}
+			else
+			{
+				gate.ptr = Identity;
+			}
+		}
+		else if (queueVector == NULL)
+		{
+			gate.ptr = gate_ptr;
+		}
+		else
+		{
+			gate.ptr = (*queueVector)[i].ptr;
+		}
+
+		tmp.rows = filter.rows * gate.rows;
+		tmp.cols = filter.cols * gate.cols;
+		tmp.ptr = malloc((size_t)tmp.rows * ((size_t)tmp.cols * (size_t)2) * (size_t)sizeof(float));
+		if (tmp.ptr == NULL)
+		{
+			fprintf(stderr, "QAnsel: Ran out of memory.\n");
+			exit(1);
+		}
+		#ifdef SPEED_TEST
+		printf("(%ix%i);(%ix%i) (knk)\n", tmp.rows, tmp.cols, gate.rows, gate.cols);
+		unsigned long int us1, us2;
+		us1 = get_time();
+		cpx_mtx_knk_metal(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tMetal: %lu\n", us2 - us1);
+		us1 = get_time();
+		cpx_mtx_knk_metal_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tMetal2x2: %lu\n", us2 - us1);
+		us1 = get_time();
+		cpx_mtx_knk_threads(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tThreads: %lu\n", us2 - us1);
+		us1 = get_time();
+		cpx_mtx_knk_threads_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tThreads2x2: %lu\n", us2 - us1);
+		us1 = get_time();
+		cpx_mtx_knk(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tBare: %lu\n", us2 - us1);
+		us1 = get_time();
+		cpx_mtx_knk_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		us2 = get_time();
+		printf("\tBare2x2: %lu\n", us2 - us1);
+
+		//us1 = get_time();
+		//cpx_mtx_knk(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		//us2 = get_time();
+		//printf("\tTranspose: %lu\n", us2 - us1);
+		#else
+		if (((ctx->optimization_level) & QANSEL_MODE_METAL) && !((ctx->optimization_level) & QANSEL_MODE_THREADED) && tmp.cols >= 64)
+		{
+			cpx_mtx_knk_metal_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		}
+		else if (((ctx->optimization_level) & QANSEL_MODE_THREADED) && tmp.cols >= 64)
+		{
+			cpx_mtx_knk_threads_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		}
+		else
+		{
+			cpx_mtx_knk_2x2(tmp.ptr, filter.ptr, gate.ptr, filter.rows, filter.cols, gate.rows, gate.cols);
+		}
+		#endif
+
+		free(filter.ptr);
+		filter.ptr = tmp.ptr;
+		filter.rows = tmp.rows;
+		filter.cols = tmp.cols;
+	}
+
+	cpx_mtx_init(&tmp, stateVector->rows, stateVector->cols);
+	
+	#ifdef SPEED_TEST
+	printf("%ix%i (dot)\n", tmp.rows, tmp.cols);
+	unsigned long int us1, us2;
+	us1 = get_time();
+	cpx_mtx_dot_metal(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	us2 = get_time();
+	printf("\tMetal: %lu\n", us2 - us1);
+	us1 = get_time();
+	cpx_mtx_dot_threads(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	us2 = get_time();
+	printf("\tThreads: %lu\n", us2 - us1);
+	us1 = get_time();
+	cpx_mtx_dot(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	us2 = get_time();
+	printf("\tBare: %lu\n", us2 - us1);
+	#else
+	if (((ctx->optimization_level) & QANSEL_MODE_METAL) && tmp.cols >= 64)
+	{
+		cpx_mtx_dot_metal(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	}
+	else if (((ctx->optimization_level) & QANSEL_MODE_THREADED) && tmp.cols >= 64)
+	{
+		cpx_mtx_dot_threads(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	}
+	else
+	{
+		cpx_mtx_dot(tmp.ptr, stateVector->ptr, filter.ptr, stateVector->rows, stateVector->cols, filter.rows, filter.cols);
+	}
+	#endif
+	free(stateVector->ptr);
+	stateVector->ptr = tmp.ptr;
+	free(filter.ptr);
+	if (needToFreeGate) free(gate_ptr);
+}
+
+unsigned char qansel_measure(QAnselContext* ctx, cpx_mtx_t* stateVector, unsigned char qubitCount, unsigned char qubit)
+{
+	unsigned int qubitCountPow2 = (unsigned int)pow(2, qubitCount);
+	cpx_t n;
+	float prob0 = 0;
+	for (unsigned int i = 0; i < qubitCountPow2; i++)
+	{
+		unsigned char bit = (i >> qubit) & 1;
+		cpx_mtx_get(stateVector, 0, i, &n);
+		if (bit == 0) prob0 += cpx_magsqr(&n);
+	}
+	float r = (ctx->hidden_variable_set) ? qansel_rand_h() : qansel_rand_t(ctx);
+	//add any error skew if desired
+	if (ctx->noise > 0)
+	{
+		float probtot;
+		float prob1 = 1 - prob0;
+		prob0 += ctx->noise;
+		prob1 += ctx->noise;
+		probtot = prob0 + prob1;
+		prob0 /= probtot;
+		prob1 /= probtot;
+	}
+	unsigned char newBit = r < prob0 ? 0 : 1;
+	float probTot = 0;
+	for (unsigned int i = 0; i < qubitCountPow2; i++)
+	{
+		unsigned char bit = (i >> qubit) & 1;
+		cpx_mtx_get(stateVector, 0, i, &n);
+		if (bit != newBit)
+		{
+			n.real = 0;
+			n.imaginary = 0;
+		}
+		else
+		{
+			probTot += cpx_magsqr(&n);
+		}
+		cpx_mtx_set(stateVector, 0, i, &n);
+	}
+	
+	float multiplier = sqrt(1 / probTot);
+	for (unsigned int i = 0; i < qubitCountPow2; i++)
+	{
+		unsigned char bit = (i >> qubit) & 1;
+		cpx_mtx_get(stateVector, 0, i, &n);
+		if (bit == newBit)
+		{
+			n.real *= multiplier;
+			n.imaginary *= multiplier;
+		}
+		cpx_mtx_set(stateVector, 0, i, &n);
+	}
+
+	return newBit;
+}
+
+int qansel_get_instruction_bitmax(unsigned char* ptr, int offset, int* bitmax, int* qbitmax)
+{
+	unsigned char a0, a1, a2;
+	*bitmax = 0;
+	*qbitmax = 0;
+	switch (ptr[offset])
+	{
+		case QANSEL_INSTRUCTION_X:
+		case QANSEL_INSTRUCTION_Y:
+		case QANSEL_INSTRUCTION_Z:
+		case QANSEL_INSTRUCTION_H:
+		case QANSEL_INSTRUCTION_S:
+		case QANSEL_INSTRUCTION_T:
+		case QANSEL_INSTRUCTION_SDG:
+		case QANSEL_INSTRUCTION_TDG:
+		case QANSEL_INSTRUCTION_RX:
+		case QANSEL_INSTRUCTION_RY:
+		case QANSEL_INSTRUCTION_RZ:
+		case QANSEL_INSTRUCTION_U1:
+		case QANSEL_INSTRUCTION_U2:
+		case QANSEL_INSTRUCTION_U3:
+		case QANSEL_INSTRUCTION_DENSITY:
+	 		a0 = ptr[offset + 1];
+			if (a0 > QANSEL_QBOUND_UPPER) return 0;
+			*qbitmax = a0 + 1;
+			return 1;
+		case QANSEL_INSTRUCTION_BARRIER:
+		case QANSEL_INSTRUCTION_BORN:
+		case QANSEL_INSTRUCTION_BSAMPLE:
+	 		a0 = ptr[offset + 1];
+			if (a0 > QANSEL_QBOUND_UPPER && a0 != QANSEL_ALL_QUANTUM) return 0;
+			if (a0 != QANSEL_ALL_QUANTUM) *qbitmax = a0 + 1;
+			return 1;
+		case QANSEL_INSTRUCTION_IF_E:
+		case QANSEL_INSTRUCTION_IF_NE:
+		case QANSEL_INSTRUCTION_IF_G:
+		case QANSEL_INSTRUCTION_IF_GE:
+		case QANSEL_INSTRUCTION_IF_L:
+		case QANSEL_INSTRUCTION_IF_LE:
+	 		a0 = ptr[offset + 1];
+			if ((a0 > QANSEL_CBOUND_UPPER || a0 < QANSEL_CBOUND_LOWER) && a0 != QANSEL_ALL_CLASSIC) return 0;
+			if (a0 != QANSEL_ALL_CLASSIC) *bitmax = (a0 - QANSEL_CBOUND_LOWER) + 1;
+			return 1;
+		case QANSEL_INSTRUCTION_SET:
+		case QANSEL_INSTRUCTION_RESET:
+		case QANSEL_INSTRUCTION_PRINT:
+			a0 = ptr[offset + 1];
+			if (a0 == QANSEL_ALL || a0 == QANSEL_ALL_QUANTUM || a0 == QANSEL_ALL_CLASSIC) return 1;
+			if (a0 <= QANSEL_QBOUND_UPPER)
+			{
+				*qbitmax = a0 + 1;
+				return 1;
+			}
+			else if (a0 >= QANSEL_CBOUND_LOWER && a0 <= QANSEL_CBOUND_UPPER)
+			{
+				*bitmax = (a0 - QANSEL_CBOUND_LOWER) + 1;
+				return 1;
+			}
+			return 0;
+		case QANSEL_INSTRUCTION_CX:
+		case QANSEL_INSTRUCTION_SWAP:
+			a0 = ptr[offset + 1];
+			a1 = ptr[offset + 2];
+			if (a0 > QANSEL_QBOUND_UPPER) return 0;
+			if (a1 > QANSEL_QBOUND_UPPER) return 0;
+			*qbitmax = (a0 > a1 ? a0 : a1) + 1;
+			return 1;
+		case QANSEL_INSTRUCTION_CCX:
+		case QANSEL_INSTRUCTION_CSWAP:
+			a0 = ptr[offset + 1];
+			a1 = ptr[offset + 2];
+			a2 = ptr[offset + 3];
+			if (a0 > QANSEL_QBOUND_UPPER || a1 > QANSEL_QBOUND_UPPER || a2 > QANSEL_QBOUND_UPPER) return 0;
+			*qbitmax = ((a0 > a1) && (a0 > a2) ? a0 : ((a1 > a0) && (a1 > a2) ? a1 : a2)) + 1;
+			return 1; 
+		case QANSEL_INSTRUCTION_MEASURE:
+			a0 = ptr[offset + 1];
+			a1 = ptr[offset + 2];
+			if (a0 > QANSEL_QBOUND_UPPER) return 0;
+			if (a1 > QANSEL_CBOUND_UPPER || a1 < QANSEL_CBOUND_LOWER) return 0;
+			*qbitmax = a0 + 1;
+			*bitmax = (a1 - QANSEL_CBOUND_LOWER) + 1;
+			return 1;
+		case QANSEL_INSTRUCTION_EXIT:
+			return 1;
+	}
+	return 0;
+}
+
+int qansel_get_instruction_size(unsigned char instr)
+{
+	switch (instr)
+	{
+		case QANSEL_INSTRUCTION_X: return 1 + 1;
+		case QANSEL_INSTRUCTION_Y: return 1 + 1;
+		case QANSEL_INSTRUCTION_Z: return 1 + 1;
+		case QANSEL_INSTRUCTION_H: return 1 + 1;
+		case QANSEL_INSTRUCTION_S: return 1 + 1;
+		case QANSEL_INSTRUCTION_T: return 1 + 1;
+		case QANSEL_INSTRUCTION_SDG: return 1 + 1;
+		case QANSEL_INSTRUCTION_TDG: return 1 + 1;
+		case QANSEL_INSTRUCTION_RX: return 1 + 1 + sizeof(float);
+		case QANSEL_INSTRUCTION_RY: return 1 + 1 + sizeof(float);
+		case QANSEL_INSTRUCTION_RZ: return 1 + 1 + sizeof(float);
+		case QANSEL_INSTRUCTION_U1: return 1 + 1 + sizeof(float);
+		case QANSEL_INSTRUCTION_U2: return 1 + 1 + sizeof(float) * 2;
+		case QANSEL_INSTRUCTION_U3: return 1 + 1 + sizeof(float) * 3;
+		case QANSEL_INSTRUCTION_CX: return 1 + 2;
+		case QANSEL_INSTRUCTION_SWAP: return 1 + 2;
+		case QANSEL_INSTRUCTION_CCX: return 1 + 3;
+		case QANSEL_INSTRUCTION_CSWAP: return 1 + 3;
+		case QANSEL_INSTRUCTION_MEASURE: return 1 + 2;
+		case QANSEL_INSTRUCTION_DENSITY: return 1 + 1;
+		case QANSEL_INSTRUCTION_BSAMPLE: return 1 + 1;
+		case QANSEL_INSTRUCTION_BORN: return 1 + 1;
+		case QANSEL_INSTRUCTION_IF_E: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_IF_NE: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_IF_G: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_IF_GE: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_IF_L: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_IF_LE: return 1 + 1 + sizeof(unsigned short);
+		case QANSEL_INSTRUCTION_SET: return 1 + 1;
+		case QANSEL_INSTRUCTION_RESET: return 1 + 1;
+		case QANSEL_INSTRUCTION_PRINT: return 1 + 1;
+		case QANSEL_INSTRUCTION_BARRIER: return 1 + 1;
+		case QANSEL_INSTRUCTION_EXIT: return 1;
+	}
+	return 0;
+}
+
+void qansel_get_barrier(QBytecode** qbc, int idx)
+{
+	switch ((*qbc)[idx].op)
+	{
+		case QANSEL_INSTRUCTION_EXIT:
+		case QANSEL_INSTRUCTION_IF_E:
+		case QANSEL_INSTRUCTION_IF_NE:
+		case QANSEL_INSTRUCTION_IF_G:
+		case QANSEL_INSTRUCTION_IF_GE:
+		case QANSEL_INSTRUCTION_IF_L:
+		case QANSEL_INSTRUCTION_IF_LE:
+			(*qbc)[idx].barrier_width = 0;
+			return;
+		case QANSEL_INSTRUCTION_X:
+		case QANSEL_INSTRUCTION_Y:
+		case QANSEL_INSTRUCTION_Z:
+		case QANSEL_INSTRUCTION_H:
+		case QANSEL_INSTRUCTION_S:
+		case QANSEL_INSTRUCTION_T:
+		case QANSEL_INSTRUCTION_SDG:
+		case QANSEL_INSTRUCTION_TDG:
+		case QANSEL_INSTRUCTION_RX:
+		case QANSEL_INSTRUCTION_RY:
+		case QANSEL_INSTRUCTION_RZ:
+		case QANSEL_INSTRUCTION_U1:
+		case QANSEL_INSTRUCTION_U2:
+		case QANSEL_INSTRUCTION_U3:
+			(*qbc)[idx].barrier_width = 1;
+			(*qbc)[idx].barrier[0] = (*qbc)[idx].bytes[1];
+			return;
+		case QANSEL_INSTRUCTION_MEASURE:
+		case QANSEL_INSTRUCTION_DENSITY:
+			(*qbc)[idx].barrier_width = 2;
+			(*qbc)[idx].barrier[0] = (*qbc)[idx].bytes[1];
+			(*qbc)[idx].barrier[1] = (*qbc)[idx].bytes[2];
+			return;
+		case QANSEL_INSTRUCTION_BSAMPLE:
+		case QANSEL_INSTRUCTION_BORN:
+		case QANSEL_INSTRUCTION_BARRIER:
+		case QANSEL_INSTRUCTION_PRINT:
+		case QANSEL_INSTRUCTION_SET:
+		case QANSEL_INSTRUCTION_RESET:
+			switch ((*qbc)[idx].bytes[1])
+			{
+				case QANSEL_ALL:
+					(*qbc)[idx].barrier_width = QANSEL_QUBITS_MAX * 2;
+					for (int i = 0; i < QANSEL_QUBITS_MAX; i++) (*qbc)[idx].barrier[i] = i;
+					for (int i = 0; i < QANSEL_QUBITS_MAX; i++) (*qbc)[idx].barrier[i] = i + QANSEL_CBOUND_LOWER;
+					return;
+				case QANSEL_ALL_QUANTUM:
+					(*qbc)[idx].barrier_width = QANSEL_QUBITS_MAX;
+					for (int i = 0; i < QANSEL_QUBITS_MAX; i++) (*qbc)[idx].barrier[i] = i;
+					return;
+				case QANSEL_ALL_CLASSIC:
+					(*qbc)[idx].barrier_width = QANSEL_QUBITS_MAX;
+					for (int i = 0; i < QANSEL_QUBITS_MAX; i++) (*qbc)[idx].barrier[i] = i + QANSEL_CBOUND_LOWER;
+					return;
+				default:
+					(*qbc)[idx].barrier_width = 1;
+					(*qbc)[idx].barrier[0] = (*qbc)[idx].bytes[1];
+					return;
+			}
+		case QANSEL_INSTRUCTION_CX:
+		case QANSEL_INSTRUCTION_SWAP:
+			(*qbc)[idx].barrier_width = 2;
+			(*qbc)[idx].barrier[0] = (*qbc)[idx].bytes[1];
+			(*qbc)[idx].barrier[1] = (*qbc)[idx].bytes[2];
+			return;
+		case QANSEL_INSTRUCTION_CCX:
+		case QANSEL_INSTRUCTION_CSWAP:
+			(*qbc)[idx].barrier_width = 3;
+			(*qbc)[idx].barrier[0] = (*qbc)[idx].bytes[1];
+			(*qbc)[idx].barrier[1] = (*qbc)[idx].bytes[2];
+			(*qbc)[idx].barrier[2] = (*qbc)[idx].bytes[3];
+			return;
+	}
+	fprintf(stderr, "QAnsel (#%i): Unknown error in barrier analysis.\n", idx);
+	exit(1);
+}
+
+void qansel_born(QAnselContext* ctx, cpx_mtx_t* stateVector, int PC, int qubitCount, unsigned char q0, unsigned char mode)
+{
+	unsigned int qubitCountPow2 = (unsigned int)pow(2, qubitCount);
+	if (q0 == QANSEL_ALL_QUANTUM)
+	{
+		float *psisquared = malloc(sizeof(float) * qubitCountPow2);
+		for (unsigned int j = 0; j < qubitCountPow2; j++)
+		{
+			cpx_t n;
+			cpx_mtx_get(stateVector, 0, j, &n);
+			psisquared[j] = cpx_magsqr(&n);
+		}
+		//add any error skew if desired
+		if (ctx->noise > 0)
+		{
+			float totalprob = 0;
+			for (unsigned int j = 0; j < qubitCountPow2; j++)
+			{
+				psisquared[j] += ctx->noise;
+				totalprob += psisquared[j];
+			}
+			for (unsigned int j = 0; j < qubitCountPow2; j++)
+			{
+				psisquared[j] /= totalprob;
+			}
+		}
+		if (mode == 0)
+		{
+			for (unsigned int j = 0; j < qubitCountPow2; j++)
+			{
+				unsigned int tmp = j;
+				for (unsigned char k = 0; k < qubitCount; k++)
+				{
+					putchar('0' + (tmp >> (qubitCount - 1) & 1));
+					tmp <<= 1;
+				}
+				printf("\t%.2f%%\n", psisquared[j] * 100);
+			}
+			free(psisquared);
+			return;
+		}
+		float *psisquared_sorted = malloc(sizeof(float) * qubitCountPow2);
+		unsigned short *sorting = malloc(sizeof(unsigned short) * qubitCountPow2);
+		unsigned short *rsorting = malloc(sizeof(unsigned short) * qubitCountPow2);
+		for (int i = 0; i < qubitCountPow2; i++) psisquared_sorted[i] = 0;
+		for (int i = 0; i < qubitCountPow2; i++)
+		{
+			float max = -1;
+			int maxi = -1;
+			for (int j = 0; j < qubitCountPow2; j++)
+			{
+				if (psisquared[j] >= max)
+				{
+					max = psisquared[j];
+					maxi = j;
+				}
+			}
+			psisquared_sorted[i] = max;
+			sorting[i] = maxi;
+			rsorting[maxi] = i;
+			psisquared[maxi] = -1;
+		}
+		//for (int i = 0; i < qubitCountPow2; i++) printf("]%i -> %i: %f[\n", rsorting[i], sorting[i], psisquared_sorted[i]);
+		free(psisquared);
+		unsigned short *stats = malloc(sizeof(unsigned short) * qubitCountPow2);
+		for (int i = 0; i < qubitCountPow2; i++) stats[i] = 0;
+		for (int i = 0; i < (ctx->bsampling_shots); i++)
+		{
+			float r = (ctx->hidden_variable_set) ? qansel_rand_h() : qansel_rand_t(ctx);
+			float j = 0;
+			//printf("--------------------------------\n");
+			for (unsigned int j = 0; j < qubitCountPow2; j++)
+			{
+				//printf("%f >= %f\n", psisquared_sorted[j], r);
+				if (psisquared_sorted[j] >= (r - 0.0000001))
+				{
+					stats[j]++;
+					//printf("%i _/\n", j);
+					break;
+				}
+				r -= psisquared_sorted[j];
+				if (j == qubitCountPow2 - 1)
+				{
+					stats[j]++;
+					//printf("%i _/\n", j);
+				}
+			}
+		}
+		for (unsigned int j = 0; j < qubitCountPow2; j++)
+		{
+			unsigned int tmp = j;
+			for (unsigned char k = 0; k < qubitCount; k++)
+			{
+				putchar('0' + (tmp >> (qubitCount - 1) & 1));
+				tmp <<= 1;
+			}
+			//printf("%i -> %i\n", j, sorting[j]);
+			printf("\t%i\t%.2f%%\n", stats[rsorting[j]], ((float)stats[rsorting[j]] / (float)(ctx->bsampling_shots)) * (float)100);
+		}
+		free(psisquared_sorted);
+		free(stats);
+		free(sorting);
+		free(rsorting);
+	}
+	else if (q0 <= QANSEL_QBOUND_UPPER)
+	{
+		float prob = 0;
+		for (unsigned int j = 0; j < qubitCountPow2; j++)
+		{
+			cpx_t n;
+			cpx_mtx_get(stateVector, 0, j, &n);
+			if ((j >> q0) & 1)
+			{
+				prob += cpx_magsqr(&n);
+			}
+		}
+		if (mode == 0)
+		{
+			printf("0\t%.2f%%\n", (1 - prob) * 100.0);
+			printf("1\t%.2f%%\n", prob * 100.0);
+			return;
+		}
+
+		unsigned short stats = 0;
+		for (int i = 0; i < (ctx->bsampling_shots); i++)
+		{
+			float r = (ctx->hidden_variable_set) ? qansel_rand_h() : qansel_rand_t(ctx);
+			stats += r < prob ? 1 : 0;
+		}
+		printf("0\t%i\t%.2f%%\n", (ctx->bsampling_shots) - stats, (((float)(ctx->bsampling_shots) - (float)stats) / (float)(ctx->bsampling_shots)) * (float)100);
+		printf("1\t%i\t%.2f%%\n", stats, ((float)stats / (float)(ctx->bsampling_shots)) * (float)100);
+	}
+}
+
+void qansel_density_or_print(cpx_mtx_t* stateVector, unsigned char* bitVector, unsigned char density, int bitCount, int qubitCount, unsigned char a0)
+{
+	unsigned int qubitCountPow2 = (unsigned int)pow(2, qubitCount);
+	if (a0 == QANSEL_ALL_QUANTUM || a0 == QANSEL_ALL_CLASSIC || a0 == QANSEL_ALL)
+	{
+		if (a0 == QANSEL_ALL_QUANTUM || a0 == QANSEL_ALL)
+		{
+			printf("[ "); cpx_mtx_print(stateVector); printf(" ]\n");
+		}
+		if (a0 == QANSEL_ALL_CLASSIC || a0 == QANSEL_ALL)
+		{
+			for (int32_t j = bitCount - 1; j >= 0; j--)
+			{
+				putchar('0' + bitVector[j]);
+			}
+			putchar('\n');
+		}
+	}
+	else if (a0 >= QANSEL_CBOUND_LOWER && a0 <= QANSEL_CBOUND_UPPER)
+	{
+		putchar('0' + bitVector[a0 - QANSEL_CBOUND_LOWER]);
+		putchar('\n');
+	}
+	else
+	{
+		cpx_mtx_t tmp;
+		cpx_mtx_init(&tmp, 1, 2);
+		for (unsigned int j = 0; j < qubitCountPow2; j++)
+		{
+			if ((j >> a0) & 1)
+			{
+				cpx_t a, b;
+				cpx_mtx_get(&tmp, 0, 1, &a);
+				cpx_mtx_get(stateVector, 0, j, &b);
+				a.real += b.real;
+				a.imaginary += b.imaginary;
+				cpx_mtx_set(&tmp, 0, 1, &a);
+			}
+			else
+			{
+				cpx_t a, b;
+				cpx_mtx_get(&tmp, 0, 0, &a);
+				cpx_mtx_get(stateVector, 0, j, &b);
+				a.real += b.real;
+				a.imaginary += b.imaginary;
+				cpx_mtx_set(&tmp, 0, 0, &a);
+			}
+		}
+		float multiplier = 0;
+		cpx_t n;
+		cpx_mtx_get(&tmp, 0, 0, &n);
+		multiplier += cpx_magsqr(&n);
+		cpx_mtx_get(&tmp, 0, 1, &n);
+		multiplier += cpx_magsqr(&n);
+		multiplier = sqrt(1 / multiplier);
+		n.real *= multiplier;
+		n.imaginary *= multiplier;
+		cpx_mtx_set(&tmp, 0, 1, &n);
+		cpx_mtx_get(&tmp, 0, 0, &n);
+		n.real *= multiplier;
+		n.imaginary *= multiplier;
+		cpx_mtx_set(&tmp, 0, 0, &n);
+
+		if (density)
+		{
+			cpx_t a, b, c, d, x, y, z, w;
+			cpx_mtx_get(&tmp, 0, 0, &a);
+			cpx_mtx_get(&tmp, 0, 1, &b);
+			cpx_mtx_get(&tmp, 0, 0, &c);
+			cpx_mtx_get(&tmp, 0, 1, &d);
+			c.imaginary *= -1;
+			d.imaginary *= -1; 
+			cpx_mul(&x, &a, &c);
+			cpx_mul(&y, &a, &d);
+			cpx_mul(&z, &b, &c);
+			cpx_mul(&w, &b, &d);
+			char* sx = cpx_str(&x);
+			char* sy = cpx_str(&y);
+			char* sz = cpx_str(&z);
+			char* sw = cpx_str(&w);
+			printf("[ %s, %s ]\n", sx, sy);
+			printf("[ %s, %s ]\n", sz, sw);
+			free(sx);
+			free(sy);
+			free(sz);
+			free(sw);
+		}
+		else
+		{
+			printf("[ "); cpx_mtx_print(&tmp); printf(" ]\n");
+		}
+		cpx_mtx_free(&tmp);
+	}
+}
+
+float qansel_get_float(unsigned char* program, int offset)
+{
+	float ret;
+	memcpy(&ret, program + offset, sizeof(float));
+	return ret;
+}
+
+short qansel_get_short(unsigned char* program, int offset)
+{
+	short ret;
+	memcpy(&ret, program + offset, sizeof(short));
+	return ret;
+}
+
+int qansel_get_int(unsigned char* program, int offset)
+{
+	int ret;
+	memcpy(&ret, program + offset, sizeof(int));
+	return ret;
+}
+
+void qansel_set(QAnselContext* ctx, cpx_mtx_t* stateVector, unsigned char* bitVector, int qubitCount, int bitCount, unsigned char q0, unsigned char value)
+{
+	unsigned int qubitCountPow2 = (unsigned int)pow(2, qubitCount);
+	if (q0 == QANSEL_ALL)
+	{
+		cpx_mtx_set2(stateVector, 0, 0, 1, 0);
+		for (unsigned int j = 1; j < qubitCountPow2; j++)
+		{
+			cpx_mtx_set2(stateVector, 0, j, 0, 0);
+		}
+		for (unsigned char j = 0; j < bitCount; j++)
+		{
+			bitVector[j] = 0;
+		}
+	}
+	else if (q0 == QANSEL_ALL_QUANTUM)
+	{
+		cpx_mtx_set2(stateVector, 0, 0, 1, 0);
+		for (unsigned int j = 1; j < qubitCountPow2; j++)
+		{
+			cpx_mtx_set2(stateVector, 0, j, 0, 0);
+		}
+	}
+	else if (q0 == QANSEL_ALL_CLASSIC)
+	{
+		for (unsigned char j = 0; j < bitCount; j++)
+		{
+			bitVector[j] = 0;
+		}
+	}
+	else if (q0 <= QANSEL_QBOUND_UPPER)
+	{
+		unsigned char bit = qansel_measure(ctx, stateVector, qubitCount, q0);
+		if (bit != value)
+		{
+			qansel_instruction(ctx, stateVector, qubitCount, QANSEL_INSTRUCTION_X, q0, 0, 0, 0, NULL);
+		}
+	}
+	else if (q0 >= QANSEL_CBOUND_LOWER && q0 <= QANSEL_CBOUND_UPPER)
+	{
+		bitVector[q0 - QANSEL_CBOUND_LOWER] = 0;
+	}
+}
+
+unsigned char qansel_compare(unsigned char* bitVector, int bitCount, int PC, unsigned char a0, short op)
+{
+	unsigned char ret = 0;
+	short val;
+	if (a0 == QANSEL_ALL_CLASSIC)
+	{
+		val = 0;
+		for (int32_t j = bitCount - 1; j >= 0; j--)
+		{
+			val = (val << 1) | bitVector[j];
+		}
+		if (val == op) ret |= QANSEL_FLAGS_EQUAL;
+		if (val > op) ret |= QANSEL_FLAGS_GREATER;
+		if (val < op) ret |= QANSEL_FLAGS_LESSER;
+	}
+	else if (a0 >= QANSEL_CBOUND_LOWER && a0 <= QANSEL_CBOUND_UPPER)
+	{
+		val = bitVector[a0 - QANSEL_CBOUND_LOWER];
+		if (val == op) ret |= QANSEL_FLAGS_EQUAL;
+		if (val > op) ret |= QANSEL_FLAGS_GREATER;
+		if (val < op) ret |= QANSEL_FLAGS_LESSER;
+	}
+	return ret;
+}
+
+//computes program efficiency
+//	points are awarded for how long strings of single qubit
+//	instructions are 
+int qansel_efficiency(QBytecode* program, int programSize, int head)
+{
+	int score = 0;
+	int tmpscore = 0;
+	int idx = head;
+
+	/*
+	printf("--------------------------------\n");
+	for (int i = 0; i < programSize; i++)
+	{
+		if (i != -1)
+		{
+			printf("%02X: %s\n", i, qansel_instruction_to_string(program[i].op));
+		}
+		else
+		{
+			printf("%02X: ERR\n", i);
+		}
+	}
+	printf("================================\n");
+	printf("%i\n", head);
+	for (int i = 0; i < programSize; i++)
+	{
+		printf("%i -> %i\n", i, program[i].next);
+	}
+	printf("================================\n");
+	*/
+
+	for (int i = 0; i < programSize; i++)
+	{
+		//printf("%02X: %s (%i)\n", i, qansel_instruction_to_string(program[idx].op), idx);
+
+		switch (program[idx].op)
+		{
+			case QANSEL_INSTRUCTION_X:
+			case QANSEL_INSTRUCTION_Y:
+			case QANSEL_INSTRUCTION_Z:
+			case QANSEL_INSTRUCTION_H:
+			case QANSEL_INSTRUCTION_S:
+			case QANSEL_INSTRUCTION_T:
+			case QANSEL_INSTRUCTION_SDG:
+			case QANSEL_INSTRUCTION_TDG:
+			case QANSEL_INSTRUCTION_RX:
+			case QANSEL_INSTRUCTION_RY:
+			case QANSEL_INSTRUCTION_RZ:
+			case QANSEL_INSTRUCTION_U1:
+			case QANSEL_INSTRUCTION_U2:
+			case QANSEL_INSTRUCTION_U3:
+				switch (tmpscore)
+				{
+					case 0: tmpscore = 1; break;
+					default: tmpscore *= 2;
+				}
+			break;
+			case QANSEL_INSTRUCTION_IF_E:
+			case QANSEL_INSTRUCTION_IF_NE:
+			case QANSEL_INSTRUCTION_IF_G:
+			case QANSEL_INSTRUCTION_IF_GE:
+			case QANSEL_INSTRUCTION_IF_L:
+			case QANSEL_INSTRUCTION_IF_LE:
+			break;
+			default:
+				if (tmpscore > 1) score += tmpscore;
+				tmpscore = 0;
+		}
+		int oidx = idx;
+		idx = program[idx].next;
+		if (idx == -1) break;
+	}
+	return score;
+}
+
+int qansel_crawl(QAnselContext* ctx, unsigned char* program, int programSize, int* qubitCount, int* bitCount)
+{
+	if (ctx->verbose) printf("Crawling program . . .\n");
+	int PC = 0;
+	*qubitCount = 0;
+	*bitCount = 0;
+	while (PC < programSize)
+	{
+		int next = qansel_get_instruction_size(program[PC]);
+		if (next == 0)
+		{
+			printf("QAnsel (%04X): Invalid instruction 0x%02x.\n", PC, program[PC]);
+			return 0;
+		}
+		int bitmax, qbitmax;
+		int success = qansel_get_instruction_bitmax(program, PC, &bitmax, &qbitmax);
+		if (!success)
+		{
+			fprintf(stderr, "QAnsel (%04X): Invalid index.\n", PC);
+			return 0;
+		}
+		if (bitmax > *bitCount) *bitCount = bitmax;
+		if (qbitmax > *qubitCount) *qubitCount = qbitmax;
+		PC += next;
+	}
+	if (*qubitCount > ctx->qubit_count)
+	{
+			fprintf(stderr, "QAnsel: Not enough qubits allocated.\n");
+			return 0;
+	}
+	if (*bitCount > ctx->bit_count)
+	{
+			fprintf(stderr, "QAnsel: Not enough classical bits allocated.\n");
+			return 0;
+	}
+	*qubitCount = ctx->qubit_count;
+	*bitCount = ctx->bit_count;
+	return 1;
+}
+
+void qansel_reorder(QAnselContext* ctx, unsigned char* program, int programSize)
+{
+	if (ctx->verbose) printf("Reordering . . .\n");
+
+	//break out program for easier manipulation
+	QBytecode* ramInstr = malloc(0);
+	int ramInstrLen = 0;
+	int copyifop = 0;
+	unsigned char ifop[16];
+	int PC = 0;
+
+	while (PC < programSize)
+	{
+		int next = qansel_get_instruction_size(program[PC]);
+		switch (program[PC])
+		{
+			case QANSEL_INSTRUCTION_IF_E:
+			case QANSEL_INSTRUCTION_IF_NE:
+			case QANSEL_INSTRUCTION_IF_G:
+			case QANSEL_INSTRUCTION_IF_GE:
+			case QANSEL_INSTRUCTION_IF_L:
+			case QANSEL_INSTRUCTION_IF_LE:
+				memcpy(ifop, program + PC, next);
+				copyifop = 1;
+				PC += next;
+			continue;
+		}
+		ramInstr = realloc(ramInstr, sizeof(QBytecode) * (ramInstrLen + 1));
+		ramInstr[ramInstrLen].size = next;
+		memcpy(ramInstr[ramInstrLen].bytes, program + PC, next);
+		ramInstr[ramInstrLen].op = program[PC];
+		ramInstr[ramInstrLen].use_ifop = copyifop;
+		if (copyifop)
+		{
+			memcpy(ramInstr[ramInstrLen].ifop, ifop, sizeof(ifop));
+			copyifop = 0;
+		}
+		ramInstrLen++;
+		PC += next;
+	}
+
+	for (int i = 0; i < ramInstrLen; i++)
+	{
+		if (i == 0)
+		{
+			ramInstr[0].prev = -1;
+			ramInstr[0].next = ramInstrLen > 1 ? 1 : -1;
+		}
+		else if (i == ramInstrLen - 1)
+		{
+			ramInstr[ramInstrLen - 1].prev = i - 1;
+			ramInstr[ramInstrLen - 1].next = -1;
+		}
+		else
+		{
+			ramInstr[i].prev = i - 1;
+			ramInstr[i].next = i + 1;
+		}
+		//we can only rearrange single-qubit instructions
+		//	so treat other instructions as pre-checked
+		switch (ramInstr[i].op)
+		{
+			case QANSEL_INSTRUCTION_X:
+			case QANSEL_INSTRUCTION_Y:
+			case QANSEL_INSTRUCTION_Z:
+			case QANSEL_INSTRUCTION_H:
+			case QANSEL_INSTRUCTION_S:
+			case QANSEL_INSTRUCTION_T:
+			case QANSEL_INSTRUCTION_SDG:
+			case QANSEL_INSTRUCTION_TDG:
+			case QANSEL_INSTRUCTION_RX:
+			case QANSEL_INSTRUCTION_RY:
+			case QANSEL_INSTRUCTION_RZ:
+			case QANSEL_INSTRUCTION_U1:
+			case QANSEL_INSTRUCTION_U2:
+			case QANSEL_INSTRUCTION_U3:
+				ramInstr[i].checked = 0;
+			break;
+			default: ramInstr[i].checked = 1;
+		}
+	}
+
+	for (int i = 0; i < ramInstrLen; i++)
+	{
+		qansel_get_barrier(&ramInstr, i);
+	}
+
+	QBytecode* reordered[2];
+	reordered[0] = malloc(ramInstrLen * sizeof(QBytecode));
+	reordered[1] = malloc(ramInstrLen * sizeof(QBytecode));
+	memcpy(reordered[0], ramInstr, ramInstrLen * sizeof(QBytecode));
+	memcpy(reordered[1], ramInstr, ramInstrLen * sizeof(QBytecode));
+	
+	int efficiency = qansel_efficiency(ramInstr, ramInstrLen, 0);
+	int head[2];
+	head[0] = 0;
+	head[1] = 0;
+
+	for (int reorid = 0; reorid < 2; reorid++)
+	{
+		while (1)
+		{
+			//find the next unchecked instruction
+			int found = -1;
+			int idx = head[reorid];
+			for (int i = 0; i < ramInstrLen; i++)
+			{
+				if (reordered[reorid][idx].checked == 0)
+				{
+					found = idx;
+					break;
+				}
+				idx = reordered[reorid][idx].next;
+			}
+			if (found == -1) break;
+			//check if it can be moved up
+			reordered[reorid][found].checked = 1;
+			int blocked = 0;
+			idx = reordered[reorid][idx].prev;
+			while (idx != -1)
+			{
+				for (int i = 0; i < reordered[reorid][idx].barrier_width; i++)
+				{
+					//verify no classical bit barrier
+					if (reordered[reorid][found].use_ifop)
+					{
+						switch (reordered[reorid][found].ifop[1])
+						{
+							case QANSEL_ALL_CLASSIC:
+							case QANSEL_ALL:
+								blocked = 1;
+							break;
+							default:
+								if (reordered[reorid][found].ifop[1] == reordered[reorid][idx].barrier[i])
+									blocked = 1;
+						}
+						if (blocked) break;
+					}
+					//verify no quantum bit barrier
+					if (reordered[reorid][found].bytes[1] == reordered[reorid][idx].barrier[i])
+					{
+						blocked = 1;
+						break;
+					}
+				}
+				if (blocked) break;
+				idx = reordered[reorid][idx].prev;
+			}
+			if (reordered[reorid][found].prev != idx)
+			{
+				QBytecode* tmpbc = malloc(ramInstrLen * sizeof(QBytecode));
+				memcpy(tmpbc, reordered[reorid], ramInstrLen * sizeof(QBytecode));
+				reordered[reorid][ reordered[reorid][found].prev ].next = reordered[reorid][found].next;
+				if (reordered[reorid][found].next != -1)
+				{
+					reordered[reorid][ reordered[reorid][found].next ].prev = reordered[reorid][found].prev;
+				}
+
+				if (idx == -1)
+				{
+					//idx.next.prev = found
+					//found.next = head
+					//found.prev = -1
+					//head = found
+					reordered[reorid][head[reorid]].prev = found;
+					reordered[reorid][found].next = head[reorid];
+					reordered[reorid][found].prev = -1;
+					head[reorid] = found;
+					//printf("head: %i\n", found);
+				}
+				else
+				{
+					/*
+						[idx].next -> .prev <- o
+						found
+					
+					*/
+					//idx.next.prev = found
+					//found.next = idx.next
+					//found.prev = idx
+					//idx.next = found
+					reordered[reorid][ reordered[reorid][idx].next ].prev = found;
+					reordered[reorid][found].next = reordered[reorid][idx].next;
+					reordered[reorid][found].prev = idx;
+					reordered[reorid][idx].next = found;
+				}
+				if (reorid == 1)
+				{
+					/*
+					printf("%i\n", head[reorid]);
+					for (int i = 0; i < ramInstrLen; i++)
+					{
+						printf("%i -> %i\n", i, reordered[reorid][i].next);
+					}
+					printf("===========================\n");
+					*/
+					int tmpeff = qansel_efficiency(reordered[reorid], ramInstrLen, head[reorid]);
+					head[reorid] = 0; //correct?
+					if (tmpeff >= efficiency)
+					{
+						efficiency = tmpeff;
+					}
+					else
+					{
+						memcpy(reordered[reorid], tmpbc, ramInstrLen * sizeof(QBytecode));
+					}
+				}
+				free(tmpbc);
+			}
+		}
+	}
+
+	efficiency = qansel_efficiency(ramInstr, ramInstrLen, 0);
+	int best = -1;
+	int besti = -1;
+	for (int i = 0; i < 2; i++)
+	{
+		int tmpeff = qansel_efficiency(reordered[i], ramInstrLen, head[i]);
+		if (tmpeff > best)
+		{
+			besti = i;
+			best = tmpeff;
+		}
+	}
+	if (best > efficiency)
+	{
+		memcpy(ramInstr, reordered[besti], ramInstrLen * sizeof(QBytecode));
+	}
+	for (int i = 0; i < 2; i++)
+	{
+		free(reordered[i]);
+	}
+
+	if (best > efficiency)
+	{
+		int copyloc = 0;
+		besti = head[besti];
+		for (int i = 0; i < ramInstrLen; i++)
+		{
+			int next;
+			if (ramInstr[besti].use_ifop)
+			{
+				next = qansel_get_instruction_size(ramInstr[besti].ifop[0]);
+				memcpy(program + copyloc, ramInstr[besti].ifop, next);
+				copyloc += next;
+			}
+			next = qansel_get_instruction_size(ramInstr[besti].op);
+			memcpy(program + copyloc, ramInstr[besti].bytes, next);
+			besti = ramInstr[besti].next;
+			copyloc += next;
+		}
+	}
+
+	if (ctx->verbose)
+	{
+		if (best > efficiency)
+		{
+			printf("Efficiency score increased from %i to %i.\n", efficiency, best);
+		}
+		else
+		{
+			printf("Could not increase efficiency.\n");
+		}
+	}
+
+	free(ramInstr);
+}
+
+void qansel_run(QAnselContext* ctx, unsigned char* program, int programSize, int qubitCount, int bitCount, unsigned char* outputBitVector)
+{
+	int useQueue = ((ctx->optimization_level) & QANSEL_MODE_SMART) ? 1 : 0;
+	int PC = 0;
+	unsigned int qubitCountPow2 = (unsigned int)pow(2, qubitCount);
+	unsigned char bitVector[bitCount];
+	memset(bitVector, 0, bitCount);
+	cpx_mtx_t stateVector;
+	cpx_mtx_t* queueVector = NULL;
+	cpx_mtx_init(&stateVector, 1, qubitCountPow2);
+	cpx_mtx_set2(&stateVector, 0, 0, 1, 0);
+	if (useQueue)
+	{
+		queueVector = malloc(qubitCount * sizeof(cpx_mtx_t));
+		for (int i = 0; i < qubitCount; i++)
+		{
+			queueVector[i].ptr = malloc(sizeof(Identity));
+			memcpy(queueVector[i].ptr, Identity, sizeof(Identity));
+			queueVector[i].rows = 2;
+			queueVector[i].cols = 2;
+		}
+	}
+	if (ctx->display_delay) { ctx->display_delay = display(&stateVector, qubitCount, ctx->display_delay); }
+	unsigned char skip = 0, a0 = 0, a1 = 0, a2 = 0;
+	unsigned char flags = 0;
+	unsigned short tmp = 0;
+	unsigned char queueFlushed = 1;
+	float f0, f1, f2;
+
+	while (PC < programSize)
+	{
+		//printf("%i;%i\n", PC, programSize);
+		int next = qansel_get_instruction_size(program[PC]);
+		if (skip)
+		{
+			skip = 0;
+		}
+		else
+		{
+			unsigned char instr = program[PC];
+
+			//flush the queue if any non-single qubit instructions are called
+			if (useQueue)
+			{
+				switch (instr)
+				{
+					case QANSEL_INSTRUCTION_X:
+					case QANSEL_INSTRUCTION_Y:
+					case QANSEL_INSTRUCTION_Z:
+					case QANSEL_INSTRUCTION_H:
+					case QANSEL_INSTRUCTION_S:
+					case QANSEL_INSTRUCTION_T:
+					case QANSEL_INSTRUCTION_SDG:
+					case QANSEL_INSTRUCTION_TDG:
+					case QANSEL_INSTRUCTION_RX:
+					case QANSEL_INSTRUCTION_RY:
+					case QANSEL_INSTRUCTION_RZ:
+					case QANSEL_INSTRUCTION_U1:
+					case QANSEL_INSTRUCTION_U2:
+					case QANSEL_INSTRUCTION_U3:
+						queueFlushed = 0;
+						break;
+					case QANSEL_INSTRUCTION_IF_E:
+					case QANSEL_INSTRUCTION_IF_NE:
+					case QANSEL_INSTRUCTION_IF_G:
+					case QANSEL_INSTRUCTION_IF_GE:
+					case QANSEL_INSTRUCTION_IF_L:
+					case QANSEL_INSTRUCTION_IF_LE:
+						break;
+					default:
+						if (!queueFlushed)
+						{
+							qansel_instruction(ctx, &stateVector, qubitCount, instr, 0, 0, 0, 0, &queueVector);
+							for (int i = 0; i < qubitCount; i++)
+							{
+								memcpy(queueVector[i].ptr, Identity, sizeof(Identity));
+							}
+							queueFlushed = 1;
+						}
+						break;
+				}
+			}
+
+			if (ctx->verbose)
+			{
+				printf("%08X: %s\n", PC, qansel_instruction_to_string(instr));
+			}
+			switch (instr)
+			{
+				case QANSEL_INSTRUCTION_X:
+				case QANSEL_INSTRUCTION_Y:
+				case QANSEL_INSTRUCTION_Z:
+				case QANSEL_INSTRUCTION_H:
+				case QANSEL_INSTRUCTION_S:
+				case QANSEL_INSTRUCTION_T:
+				case QANSEL_INSTRUCTION_SDG:
+				case QANSEL_INSTRUCTION_TDG:
+				case QANSEL_INSTRUCTION_RX:
+				case QANSEL_INSTRUCTION_RY:
+				case QANSEL_INSTRUCTION_RZ:
+				case QANSEL_INSTRUCTION_U1:
+				case QANSEL_INSTRUCTION_U2:
+				case QANSEL_INSTRUCTION_U3:
+					switch (instr)
+					{
+						case QANSEL_INSTRUCTION_RX:
+							f0 = qansel_get_float(program, PC + 2);
+							f1 = -M_PI / 2;
+							f2 = M_PI / 2;
+						break;
+						case QANSEL_INSTRUCTION_RY:
+						case QANSEL_INSTRUCTION_U1:
+							f0 = qansel_get_float(program, PC + 2);
+							f1 = 0;
+							f2 = 0;
+						break;
+						case QANSEL_INSTRUCTION_RZ:
+							f0 = 0;
+							f1 = 0;
+							f2 = qansel_get_float(program, PC + 2);
+						break;
+						case QANSEL_INSTRUCTION_U2:
+							f0 = qansel_get_float(program, PC + 2);
+							f1 = qansel_get_float(program, PC + 2 + sizeof(float));
+							f2 = 0;
+						break;
+						case QANSEL_INSTRUCTION_U3:
+							f0 = qansel_get_float(program, PC + 2);
+							f1 = qansel_get_float(program, PC + 2 + sizeof(float));
+							f2 = qansel_get_float(program, PC + 2 + sizeof(float) * 2);
+						break;
+						default:
+							f0 = 0;
+							f1 = 0;
+							f2 = 0;
+						break;
+					}
+					a0 = program[PC + 1];
+					if (useQueue) qansel_instruction_queue(&queueVector, qubitCount, instr, a0, f0, f1, f2);
+					else qansel_instruction(ctx, &stateVector, qubitCount, instr, a0, f0, f1, f2, NULL);
+				break;
+				case QANSEL_INSTRUCTION_CX:
+					a0 = program[PC + 1];
+					a1 = program[PC + 2];
+					qansel_cnot(&stateVector, qubitCount, a0, a1);
+				break;
+				case QANSEL_INSTRUCTION_SWAP:
+					a0 = program[PC + 1];
+					a1 = program[PC + 2];
+					qansel_swap(&stateVector, qubitCount, a0, a1);
+				break;
+				case QANSEL_INSTRUCTION_CCX:
+					a0 = program[PC + 1];
+					a1 = program[PC + 2];
+					a2 = program[PC + 3];
+					qansel_toffoli(&stateVector, qubitCount, a0, a1, a2);
+				break;
+				case QANSEL_INSTRUCTION_CSWAP:
+					a0 = program[PC + 1];
+					a1 = program[PC + 2];
+					a2 = program[PC + 3];
+					qansel_fredkin(&stateVector, qubitCount, a0, a1, a2);
+				break;
+				case QANSEL_INSTRUCTION_MEASURE:
+					a0 = program[PC + 1];
+					a1 = program[PC + 2] - QANSEL_CBOUND_LOWER;
+					bitVector[a1] = qansel_measure(ctx, &stateVector, qubitCount, a0);
+				break;
+				case QANSEL_INSTRUCTION_BORN:
+					a0 = program[PC + 1];
+					qansel_born(ctx, &stateVector, PC, qubitCount, a0, 0);
+				break;
+				case QANSEL_INSTRUCTION_BSAMPLE:
+					a0 = program[PC + 1];
+					qansel_born(ctx, &stateVector, PC, qubitCount, a0, 1);
+				break;
+				case QANSEL_INSTRUCTION_DENSITY:
+					a0 = program[PC + 1];
+					qansel_density_or_print(&stateVector, bitVector, 1, bitCount, qubitCount, a0);
+				break;
+				case QANSEL_INSTRUCTION_PRINT:
+					a0 = program[PC + 1];
+					qansel_density_or_print(&stateVector, bitVector, 0, bitCount, qubitCount, a0);
+				break;
+				case QANSEL_INSTRUCTION_BARRIER:
+					a0 = program[PC + 1];
+				break;
+				case QANSEL_INSTRUCTION_SET:
+					a0 = program[PC + 1];
+					qansel_set(ctx, &stateVector, bitVector, qubitCount, bitCount, a0, 1);
+				break;
+				case QANSEL_INSTRUCTION_RESET:
+					a0 = program[PC + 1];
+					qansel_set(ctx, &stateVector, bitVector, qubitCount, bitCount, a0, 0);
+				break;
+				case QANSEL_INSTRUCTION_IF_E:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if (flags & QANSEL_FLAGS_EQUAL) skip = 0;
+				break;
+				case QANSEL_INSTRUCTION_IF_NE:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if (!(flags & QANSEL_FLAGS_EQUAL)) skip = 0;
+				break;
+				case QANSEL_INSTRUCTION_IF_G:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if (flags & QANSEL_FLAGS_GREATER) skip = 0;
+				break;
+				case QANSEL_INSTRUCTION_IF_GE:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if ((flags & QANSEL_FLAGS_GREATER) && (flags & QANSEL_FLAGS_EQUAL)) skip = 0;
+				break;
+				case QANSEL_INSTRUCTION_IF_L:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if (flags & QANSEL_FLAGS_LESSER) skip = 0;
+				break;
+				case QANSEL_INSTRUCTION_IF_LE:
+					a0 = program[PC + 1];
+					tmp = qansel_get_short(program, PC + 2);
+					flags = qansel_compare(bitVector, bitCount, PC, a0, tmp);
+					skip = 1; if ((flags & QANSEL_FLAGS_LESSER) && (flags & QANSEL_FLAGS_EQUAL)) skip = 0;
+				break;
+
+			}
+		}
+		PC += next;
+		if (ctx->display_delay) { ctx->display_delay = display(&stateVector, qubitCount, ctx->display_delay); }
+	}
+	if (outputBitVector != NULL)
+	{
+		for (int i = 0; i < bitCount; i++)
+		{
+			outputBitVector[i] = bitVector[i];
+		}
+	}
+	cpx_mtx_free(&stateVector);
+	if (useQueue)
+	{
+		for (int i = 0; i < qubitCount; i++)
+		{
+			free(queueVector[i].ptr);
+		}
+		free(queueVector);
+	}
+	if (ctx->display_delay) { display(NULL, 0, 0); }
+
+}
+
+int qanselExecuteBytecode(unsigned char* buff, int sizeofbuff, QAnselContext* ctx)
+{
+	if (!qanselContextValidate(ctx)) return 0;
+	qansel_rand_s(qansel_hardware_getseed());
+
+	if (ctx != NULL && ctx->hidden_variable_set)
+	{
+		unsigned int tmp;
+		memcpy(&tmp, &(ctx->hidden_variable), sizeof(unsigned int));
+		srand(tmp);
+	}
+
+	unsigned short vals;
+	float valf; 
+	int pos = 0;
+
+	int qubitCount, bitCount;
+	if (!qansel_crawl(ctx, buff, sizeofbuff, &qubitCount, &bitCount))
+	{
+		return 0;
+	}
+
+	if (ctx != NULL && ctx->sampling_shots > 0)
+	{
+		unsigned short stats[65536];
+		for (unsigned int i = 0; i < (1 << bitCount); i++)
+		{
+			stats[i] = 0;
+		}
+		unsigned char bitVect[bitCount];
+		memset(bitVect, 0, bitCount);
+		for (int i = 0; i < bitCount; i++) bitVect[i] = 0;
+		unsigned int shots = ctx->sampling_shots;
+		for (unsigned int i = 0; i < shots; i++)
+		{
+			qansel_run(ctx, buff, sizeofbuff, qubitCount, bitCount, bitVect);
+			unsigned short stat = 0;
+			for (signed char j = bitCount - 1; j >= 0; j--)
+			{
+				stat = (stat << 1) | bitVect[j];
+			}
+			stats[stat]++;
+		}
+		unsigned int count = 0;
+		for (unsigned int i = 0; i < (1 << bitCount); i++)
+		{
+			unsigned int tmp = i;
+			for (unsigned char j = 0; j < bitCount; j++)
+			{
+				unsigned char bit = (tmp >> (bitCount - 1) & 1);
+				if (j == (bitCount - (ctx->sampling_bit) - 1) && bit)
+				{
+					count += stats[i];
+				}
+				if ((ctx->sampling_bit) == QANSEL_ALL)
+				{
+					putchar('0' + bit);
+				}
+				tmp <<= 1;
+			}
+			if ((ctx->sampling_bit) == QANSEL_ALL)
+			{
+				printf("\t%i\t%.2f%%\n", stats[i], ((float)stats[i] / (float)shots) * (float)100);
+			}
+		}
+		if ((ctx->sampling_bit) != QANSEL_ALL)
+		{
+			float prob = ((float)count / (float)shots) * (float)100;
+			printf("0\t%i\t%.2f%%\n", shots-count, ((float)100)-prob);
+			printf("1\t%i\t%.2f%%\n", count, prob);
+		}
+	}
+	else
+	{
+		qansel_run(ctx, buff, sizeofbuff, qubitCount, bitCount, NULL);
+	}
+	return 1;
+}
-- 
2.39.5