examples/braket_features/Getting_Started_with_OpenQASM_on_Braket.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "80787a2d092c6fe2",
   "metadata": {},
   "source": [
    "# Getting Started with OpenQASM on Braket"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f30932accbfc8959",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:03:29.377972Z",
     "start_time": "2023-12-04T04:03:29.350550Z"
    }
   },
   "outputs": [],
   "source": [
    "# Use Braket SDK Cost Tracking to estimate the cost to run this example\n",
    "from braket.tracking import Tracker\n",
    "\n",
    "t = Tracker().start()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a2b23cf2abc23663",
   "metadata": {},
   "source": [
    "OpenQASM is a popular human-readable and hardware-agnostic quantum circuit description language. It is open-source and has been actively maintained by a [Technical Steering Committee](https://medium.com/qiskit/introducing-a-technical-steering-committee-for-openqasm3-f9db808108e1) formed by IBM, Amazon, Microsoft and the University of Innsbruck. Amazon Braket now supports OpenQASM 3.0 as an *Intermediate Representation* (IR) in addition to the in-house *JSON-Based AWS Quantum Circuit Description* ([JAQCD](https://github.com/amazon-braket/amazon-braket-schemas-python/tree/main/src/braket/ir/jaqcd)). In this notebook, we demonstrate how to submit OpenQASM quantum tasks to various devices on Braket and introduce some OpenQASM features available on Braket."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee898521529c0469",
   "metadata": {},
   "source": [
    "## Create and submit an OpenQASM quantum task\n",
    "\n",
    "Submitting a quantum task with OpenQASM is just as simple as using JAQCD. You can use the Amazon Braket Python SDK, Boto3, or the AWS CLI to submit OpenQASM 3.0 quantum tasks to an Amazon Braket device. We will go over each method in this section.\n",
    "\n",
    "\n",
    "### A Bell state\n",
    "\n",
    "We will start with by preparing a [Bell state](https://en.wikipedia.org/wiki/Bell_state) in OpenQASM:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ca2bc5f0a29e6dd6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:03:31.815079Z",
     "start_time": "2023-12-04T04:03:31.796365Z"
    }
   },
   "outputs": [],
   "source": [
    "bell_qasm = \"\"\"\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit[2] q;\n",
    "bit[2] c;\n",
    "\n",
    "h q[0];\n",
    "cnot q[0], q[1];\n",
    "\n",
    "c = measure q;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9b588c2b02979f6",
   "metadata": {},
   "source": [
    "Compare this to the same Bell state written in JAQCD:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a378dc21e3ede3d0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:03:32.662932Z",
     "start_time": "2023-12-04T04:03:32.608503Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"braketSchemaHeader\": {\n",
      "    \"name\": \"braket.ir.jaqcd.program\",\n",
      "    \"version\": \"1\"\n",
      "  },\n",
      "  \"instructions\": [\n",
      "    {\n",
      "      \"target\": 0,\n",
      "      \"type\": \"h\"\n",
      "    },\n",
      "    {\n",
      "      \"control\": 0,\n",
      "      \"target\": 1,\n",
      "      \"type\": \"cnot\"\n",
      "    }\n",
      "  ],\n",
      "  \"results\": null,\n",
      "  \"basis_rotation_instructions\": null\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "from braket.ir.jaqcd import CNot, H, Program\n",
    "\n",
    "program = Program(instructions=[H(target=0), CNot(control=0, target=1)])\n",
    "print(program.json(indent=2))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8f68befdae02c0",
   "metadata": {},
   "source": [
    "Immediately, we can see a difference: In OpenQASM, users define their own qubit registers, and thus the syntax is closer to what quantum algorithm researchers are used to; on the other hand, in JAQCD, qubits are indexed by integers and the convention is closer to that of hardware providers. Also, JAQCD has result types and basis rotation instructions embedded in the language while OpenQASM doesn't support them inherently (but later we will show how to use the `pragma` syntax to support them in OpenQASM)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a92df8c76a81db2c",
   "metadata": {},
   "source": [
    "\n",
    "### Use the Python SDK to create OpenQASM 3.0 quantum tasks\n",
    "\n",
    "Most Braket users might want to use the Braket Python SDK to submit OpenQASM quantum tasks. To submit our Bell state program in the Python SDK, we first choose the quantum device that we want to run our  program on. In this example, we will use the SV1 state-vector simulator for demonstration."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "8b433a601a62b11d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:03:38.321971Z",
     "start_time": "2023-12-04T04:03:36.978785Z"
    }
   },
   "outputs": [],
   "source": [
    "from braket.aws import AwsDevice\n",
    "from braket.devices import Devices\n",
    "\n",
    "sv1 = AwsDevice(Devices.Amazon.SV1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d46340ec242cfc8b",
   "metadata": {},
   "source": [
    "To submit the OpenQASM quantum task, we initialize an `OpenQASMProgram` object using the Bell state program text string `bell_qasm` we defined above and send it to the SV1 simulator."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1e81325c5357195d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:03:41.134199Z",
     "start_time": "2023-12-04T04:03:39.866510Z"
    },
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "from braket.ir.openqasm import Program as OpenQASMProgram\n",
    "\n",
    "bell_program = OpenQASMProgram(source=bell_qasm)\n",
    "bell_task = sv1.run(\n",
    "    bell_program,\n",
    "    shots=100,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba9729ce6ab2bf91",
   "metadata": {},
   "source": [
    "###  Submit OpenQASM 3.0 programs using the AWS Command Line Interface\n",
    "\n",
    "Alternatively, if you like the command line experience or you are not a Python user, you can also choose to use the [AWS Command Line Interface (CLI)](https://aws.amazon.com/cli/) to submit our Bell state program. Before doing that we have to make sure we have [AWS CLI installed](https://docs.aws.amazon.com/cli/latest/userguide/getting-started-install.html). The following code saves the `bell_qasm` string to a file named `bell.qasm`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "61362f489016dd1f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.407834Z",
     "start_time": "2023-11-21T08:32:41.403524Z"
    }
   },
   "outputs": [],
   "source": [
    "with open(\"bell.qasm\", \"w\") as f:\n",
    "    f.write(bell_qasm)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5c4be932b51fd2ac",
   "metadata": {},
   "source": [
    "Then we can use the command below to submit the quantum task via AWS CLI. Remember to replace the placeholder \\\"amazon-braket-my-bucket\\\" with your own bucket name.\n",
    "      \n",
    "    aws braket create-quantum-task \\\n",
    "    --region \"us-west-1\" \\\n",
    "    --device-arn \"arn:aws:braket:us-west-1::device/qpu/rigetti/Ankaa-3\" \\\n",
    "    --shots 100 \\\n",
    "    --action '{\n",
    "        \"braketSchemaHeader\": {\n",
    "            \"name\": \"braket.ir.openqasm.program\", \n",
    "            \"version\": \"1\"\n",
    "        },\n",
    "        \"source\": $(cat bell.qasm)\n",
    "    }'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68222e9240c91d58",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "## Convert OpenQASM 3.0 programs to circuits\n",
    "\n",
    "You can convert OpenQASM programs into Braket `Circuit` objects if you want to programmatically change your program."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "96417df481e2f5d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.420933Z",
     "start_time": "2023-11-21T08:32:41.407658Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T  : │  0  │  1  │  2  │\n",
      "      ┌───┐       ┌───┐ \n",
      "q0 : ─┤ H ├───●───┤ M ├─\n",
      "      └───┘   │   └───┘ \n",
      "            ┌─┴─┐ ┌───┐ \n",
      "q1 : ───────┤ X ├─┤ M ├─\n",
      "            └───┘ └───┘ \n",
      "T  : │  0  │  1  │  2  │\n"
     ]
    }
   ],
   "source": [
    "from braket.circuits import Circuit\n",
    "\n",
    "print(Circuit.from_ir(bell_qasm))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4fec44c2bb4680df",
   "metadata": {},
   "source": [
    "## Figure out what OpenQASM features are supported on each device\n",
    "\n",
    "Different devices on Braket support different subsets of OpenQASM features. To see what are the supported OpenQASM features on each device, we can simply check the device capability for OpenQASM actions. As an example, we can take a look at the `action` field in the device capability of SV1 simulator:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "78a5c9f03581cea4",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.425542Z",
     "start_time": "2023-11-21T08:32:41.422663Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ccnot',\n",
       " 'cnot',\n",
       " 'cphaseshift',\n",
       " 'cphaseshift00',\n",
       " 'cphaseshift01',\n",
       " 'cphaseshift10',\n",
       " 'cswap',\n",
       " 'cy',\n",
       " 'cz',\n",
       " 'ecr',\n",
       " 'h',\n",
       " 'i',\n",
       " 'iswap',\n",
       " 'pswap',\n",
       " 'phaseshift',\n",
       " 'rx',\n",
       " 'ry',\n",
       " 'rz',\n",
       " 's',\n",
       " 'si',\n",
       " 'swap',\n",
       " 't',\n",
       " 'ti',\n",
       " 'v',\n",
       " 'vi',\n",
       " 'x',\n",
       " 'xx',\n",
       " 'xy',\n",
       " 'y',\n",
       " 'yy',\n",
       " 'z',\n",
       " 'zz',\n",
       " 'gpi',\n",
       " 'gpi2',\n",
       " 'ms']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# supportedOperations\n",
    "sv1.properties.action[\"braket.ir.openqasm.program\"].supportedOperations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "85fbc40dc254813e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.463833Z",
     "start_time": "2023-11-21T08:32:41.425725Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['braket_unitary_matrix',\n",
       " 'braket_basis_rotation',\n",
       " 'braket_result_type_sample',\n",
       " 'braket_result_type_expectation',\n",
       " 'braket_result_type_variance',\n",
       " 'braket_result_type_probability',\n",
       " 'braket_result_type_amplitude',\n",
       " 'braket_result_type_adjoint_gradient']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# supportedPragmas\n",
    "sv1.properties.action[\"braket.ir.openqasm.program\"].supportedPragmas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1fa721d9960f5412",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.464559Z",
     "start_time": "2023-11-21T08:32:41.429030Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['braket_result_type_state_vector',\n",
       " 'braket_result_type_density_matrix',\n",
       " 'braket_noise_amplitude_damping',\n",
       " 'braket_noise_bit_flip',\n",
       " 'braket_noise_depolarizing',\n",
       " 'braket_noise_kraus',\n",
       " 'braket_noise_pauli_channel',\n",
       " 'braket_noise_generalized_amplitude_damping',\n",
       " 'braket_noise_phase_flip',\n",
       " 'braket_noise_phase_damping',\n",
       " 'braket_noise_two_qubit_dephasing',\n",
       " 'braket_noise_two_qubit_depolarizing']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# forbiddenPragmas\n",
    "sv1.properties.action[\"braket.ir.openqasm.program\"].forbiddenPragmas"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8190bb8089c7db3f",
   "metadata": {},
   "source": [
    "The SV1 OpenQASM `action` field lists supported/forbidden OpenQASM features on the device, including `supportedPragmas`, `forbiddenPragmas`, `maximumQubitArrays`, `maximumClassicalArrays`, `requiresAllQubitsMeasurement`, `supportedResultTypes`, etc. The names are self-evident, but readers are encouraged to visit the [Amazon Braket developer guide](https://docs.aws.amazon.com/braket/latest/developerguide/braket-using.html) for full information of what these fields mean."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee537a4601e4c244",
   "metadata": {},
   "source": [
    "# OpenQASM features on Braket\n",
    "\n",
    "Braket supports many useful OpenQASM features, either through the OpenQASM program syntax or Braket-specific pragmas. We will walk through some of these features in this section.\n",
    "\n",
    "## Simulating Noise with OpenQASM\n",
    "\n",
    "With the fully on-demand, high-performance, density-matrix simulator [DM1](https://docs.aws.amazon.com/braket/latest/developerguide/braket-devices.html#braket-simulator-dm1), you can easily investigate the effects of realistic noise on your quantum programs. Now, we show how to use OpenQASM programs to leverage the circuit-level noise simulation capability of DM1.\n",
    "\n",
    "To simulate noise, we have to be able to specify different noise channels. Although syntax for noise channels is not available in the OpenQASM language, Braket uses the `pragma` statement to extend OpenQASM for defining noise channels. Here is an example of an OpenQASM program that prepares a noisy 3-qubit [GHZ state](https://en.wikipedia.org/wiki/Greenberger%E2%80%93Horne%E2%80%93Zeilinger_state):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4d8a62cfc51a18b",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.464621Z",
     "start_time": "2023-11-21T08:32:41.431447Z"
    }
   },
   "outputs": [],
   "source": [
    "noisy_ghz3_program = \"\"\"\n",
    "// noisy_ghz3.qasm\n",
    "// Prepare a 3 noisy qubit GHZ state\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit[3] q;\n",
    "bit[3] c;\n",
    "\n",
    "h q[0];\n",
    "#pragma braket noise depolarizing(0.1) q[0]\n",
    "cnot q[0], q[1];\n",
    "#pragma braket noise depolarizing(0.1) q[0]\n",
    "#pragma braket noise depolarizing(0.1) q[1]\n",
    "cnot q[1], q[2];\n",
    "#pragma braket noise depolarizing(0.1) q[0]\n",
    "#pragma braket noise depolarizing(0.1) q[1]\n",
    "\n",
    "c = measure q;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a988726bb4c39e8",
   "metadata": {},
   "source": [
    "In the example above, we inserted the depolarizing noise channel with probability of 0.1 after each gate in the circuit. The `noisy_ghz3_program` is equivalent to the following program in the Braket SDK:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "647e3717ccbba672",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:41.464801Z",
     "start_time": "2023-11-21T08:32:41.434396Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Circuit('instructions': [Instruction('operator': H('qubit_count': 1), 'target': QubitSet([Qubit(0)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': Depolarizing(0.1), 'target': QubitSet([Qubit(0)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': CNot('qubit_count': 2), 'target': QubitSet([Qubit(0), Qubit(1)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': Depolarizing(0.1), 'target': QubitSet([Qubit(0)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': Depolarizing(0.1), 'target': QubitSet([Qubit(1)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': CNot('qubit_count': 2), 'target': QubitSet([Qubit(1), Qubit(2)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': Depolarizing(0.1), 'target': QubitSet([Qubit(1)]), 'control': QubitSet([]), 'control_state': (), 'power': 1), Instruction('operator': Depolarizing(0.1), 'target': QubitSet([Qubit(2)]), 'control': QubitSet([]), 'control_state': (), 'power': 1)])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from braket.circuits import Circuit, noises\n",
    "\n",
    "noisy_ghz3_circ = Circuit().h(0).cnot(0, 1).cnot(1, 2)\n",
    "noise = noises.Depolarizing(probability=0.1)\n",
    "noisy_ghz3_circ.apply_gate_noise(noise)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1db047fbac6a26bb",
   "metadata": {},
   "source": [
    "To see if `noisy_ghz3_program` and `noisy_ghz3_circ` are indeed the same, we can run both circuits and compare the results:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "4c57de26c172aa5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:42.963722Z",
     "start_time": "2023-11-21T08:32:41.441098Z"
    }
   },
   "outputs": [],
   "source": [
    "dm1 = AwsDevice(Devices.Amazon.DM1)\n",
    "\n",
    "noisy_ghz3_circ_task = dm1.run(noisy_ghz3_circ, shots=10)\n",
    "noisy_ghz3_program_task = dm1.run(OpenQASMProgram(source=noisy_ghz3_program), shots=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "75984e0e95dc4afb",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:45.959888Z",
     "start_time": "2023-11-21T08:32:42.975561Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sdk measurement results: Counter({'000': 4, '111': 3, '101': 1, '011': 1, '100': 1})\n",
      "openqasm measurement results: Counter({'111': 7, '000': 3})\n"
     ]
    }
   ],
   "source": [
    "sdk_result = noisy_ghz3_circ_task.result()\n",
    "openqasm_result = noisy_ghz3_program_task.result()\n",
    "sdk_measurement = sdk_result.measurement_counts\n",
    "openqasm_measurement = openqasm_result.measurement_counts\n",
    "print(\"sdk measurement results:\", sdk_measurement)\n",
    "print(\"openqasm measurement results:\", openqasm_measurement)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2720672ca96b104",
   "metadata": {},
   "source": [
    "As expected, the measurement counts of the two are very close.\n",
    "\n",
    "In addition to depolarizing noises, we can simulate more complicated noise types with Braket, e.g., `pauli_channel`, `amplitude_damping`, etc. Check the [Amazon Braket developer guide](https://docs.aws.amazon.com/braket/latest/developerguide/braket-using.html) for a complete list of noise channels supported on Braket. Here we give another example of general noise channels defined by the Kraus representation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "f3231a7767d20640",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:45.971122Z",
     "start_time": "2023-11-21T08:32:45.960859Z"
    }
   },
   "outputs": [],
   "source": [
    "noisy_program_with_kraus_operators = \"\"\"\n",
    "// noisy_program_with_kraus_operators\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit[2] q;\n",
    "bit[2] c;\n",
    "\n",
    "h q[0];\n",
    "#pragma braket noise kraus([[0.9486833, 0], [0, 0.9486833]], [[0, 0.31622777], [0.31622777, 0]]) q[0]\n",
    "cnot q[0], q[1];\n",
    "\n",
    "c = measure q;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ebac8bbb6d45341d",
   "metadata": {},
   "source": [
    "We inserted a single qubit noise channel defined by two 2x2 complex Kraus operators in the example above on qubit `q[0]`. Braket will validate if the Kraus operators indeed form a Completely-Positive and Trace-Preserving (CPTP) map."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ffa5e09c2ceab140",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "The `from_ir` method supports noise operations, including general Kraus operators."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "69c7ae1c396ac705",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:46.004276Z",
     "start_time": "2023-11-21T08:32:45.977541Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "noisy_ghz3_program:\n",
      "T  : │         0         │                1                │         2         │  3  │\n",
      "      ┌───┐ ┌───────────┐       ┌───────────┐ ┌───────────┐                     ┌───┐ \n",
      "q0 : ─┤ H ├─┤ DEPO(0.1) ├───●───┤ DEPO(0.1) ├─┤ DEPO(0.1) ├─────────────────────┤ M ├─\n",
      "      └───┘ └───────────┘   │   └───────────┘ └───────────┘                     └───┘ \n",
      "                          ┌─┴─┐ ┌───────────┐                     ┌───────────┐ ┌───┐ \n",
      "q1 : ─────────────────────┤ X ├─┤ DEPO(0.1) ├─────────────────●───┤ DEPO(0.1) ├─┤ M ├─\n",
      "                          └───┘ └───────────┘                 │   └───────────┘ └───┘ \n",
      "                                                            ┌─┴─┐               ┌───┐ \n",
      "q2 : ───────────────────────────────────────────────────────┤ X ├───────────────┤ M ├─\n",
      "                                                            └───┘               └───┘ \n",
      "T  : │         0         │                1                │         2         │  3  │\n",
      "\n",
      "noisy_program_with_kraus_operators:\n",
      "T  : │     0      │  1  │  2  │\n",
      "      ┌───┐ ┌────┐       ┌───┐ \n",
      "q0 : ─┤ H ├─┤ KR ├───●───┤ M ├─\n",
      "      └───┘ └────┘   │   └───┘ \n",
      "                   ┌─┴─┐ ┌───┐ \n",
      "q1 : ──────────────┤ X ├─┤ M ├─\n",
      "                   └───┘ └───┘ \n",
      "T  : │     0      │  1  │  2  │\n",
      "\n",
      "Kraus operators:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([[0.9486833+0.j, 0.       +0.j],\n",
       "        [0.       +0.j, 0.9486833+0.j]]),\n",
       " array([[0.        +0.j, 0.31622777+0.j],\n",
       "        [0.31622777+0.j, 0.        +0.j]])]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"noisy_ghz3_program:\")\n",
    "print(Circuit.from_ir(noisy_ghz3_program))\n",
    "print()\n",
    "print(\"noisy_program_with_kraus_operators:\")\n",
    "circuit_kraus = Circuit.from_ir(noisy_program_with_kraus_operators)\n",
    "print(circuit_kraus)\n",
    "print()\n",
    "print(\"Kraus operators:\")\n",
    "circuit_kraus.instructions[1].operator.to_matrix()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6b1ce4e1900550",
   "metadata": {},
   "source": [
    "## Submitting parametrized quantum tasks with OpenQASM\n",
    "\n",
    "The on-demand [SV1 simulator](https://docs.aws.amazon.com/braket/latest/developerguide/braket-devices.html#braket-simulator-sv1) and [DM1 simulator](https://docs.aws.amazon.com/braket/latest/developerguide/braket-devices.html#braket-simulator-dm1) support submitting OpenQASM programs with free parameters. You can set the value of the parameter when you submit the quantum task, like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "71245a134d32fa26",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:46.007177Z",
     "start_time": "2023-11-21T08:32:46.003030Z"
    }
   },
   "outputs": [],
   "source": [
    "parameter_qasm = \"\"\"\n",
    "OPENQASM 3.0;\n",
    "input float alpha;\n",
    "\n",
    "bit[2] b;\n",
    "qubit[2] q;\n",
    "\n",
    "h q[0];\n",
    "h q[1];\n",
    "rx(alpha) q[0];\n",
    "rx(alpha) q[1];\n",
    "b[0] = measure q[0];\n",
    "b[1] = measure q[1];\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bec130ddc2409db5",
   "metadata": {},
   "source": [
    "The `input float alpha` line indicates that we have an input parameter of type `float` named `alpha`. We can specify a value for `alpha` when we submit the quantum task using the optional `inputs` argument to `run`. `input` should be a `dict` of `string`-`float` pairs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b831353ac6cfab2a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:48.743543Z",
     "start_time": "2023-11-21T08:32:46.007096Z"
    },
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'01': 4, '10': 2, '11': 2, '00': 2})\n"
     ]
    }
   ],
   "source": [
    "input_dict = {\"alpha\": 0.1}\n",
    "param_sv1_task = sv1.run(OpenQASMProgram(source=parameter_qasm), shots=10, inputs=input_dict)\n",
    "param_sv1_result = param_sv1_task.result()\n",
    "print(param_sv1_result.measurement_counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb90a29d1debc7b4",
   "metadata": {},
   "source": [
    "Similarly, we can specify values for parametrized noise operations:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "94f2688b3380c621",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:48.746739Z",
     "start_time": "2023-11-21T08:32:48.743876Z"
    }
   },
   "outputs": [],
   "source": [
    "parameter_noise_qasm = \"\"\"\n",
    "OPENQASM 3.0;\n",
    "input float beta;\n",
    "input float alpha;\n",
    "bit[2] b;\n",
    "qubit[2] q;\n",
    "h q[0];\n",
    "h q[1];\n",
    "rx(alpha) q[0];\n",
    "rx(alpha) q[1];\n",
    "#pragma braket noise bit_flip(beta) q[0]\n",
    "b[0] = measure q[0];\n",
    "b[1] = measure q[1];\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "101d366dca84ddb2",
   "metadata": {},
   "source": [
    "We can see there are now two `input` lines, one for each free parameter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "a0aa8a773b16dc22",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:51.579287Z",
     "start_time": "2023-11-21T08:32:48.747668Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'01': 4, '10': 2, '11': 2, '00': 2})\n"
     ]
    }
   ],
   "source": [
    "noise_input_dict = {\"alpha\": 0.1, \"beta\": 0.2}\n",
    "param_dm1_task = dm1.run(OpenQASMProgram(source=parameter_qasm), shots=10, inputs=noise_input_dict)\n",
    "param_dm1_result = param_dm1_task.result()\n",
    "print(param_dm1_result.measurement_counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e60112ddc974a94",
   "metadata": {},
   "source": [
    "## Simulating arbitrary unitaries with OpenQASM\n",
    "\n",
    "The on-demand [SV1 simulator](https://docs.aws.amazon.com/braket/latest/developerguide/braket-devices.html#braket-simulator-sv1) allows us to simulate arbitrary unitary gates in a circuit. With OpenQASM, we can use the `unitary` pramga to insert these arbitrary unitary gates: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ffc32efec9315f0",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:51.589214Z",
     "start_time": "2023-11-21T08:32:51.579008Z"
    }
   },
   "outputs": [],
   "source": [
    "program_with_unitary = \"\"\"\n",
    "// noisy_program_with_kraus_operators\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit q;\n",
    "bit c;\n",
    "\n",
    "#pragma braket unitary([[0, -1im], [1im, 0]]) q\n",
    "c = measure q;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c190474a6e28563",
   "metadata": {},
   "source": [
    "The `1im` in the `unitary` pragma is the OpenQASM notation of the imaginary number $i$, thus, we were simply using the pragma to perform a Pauli Y gate. We can check it by submitting the above program to SV1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "6d687471d81b2367",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:54.390589Z",
     "start_time": "2023-11-21T08:32:51.586164Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'1': 10})\n"
     ]
    }
   ],
   "source": [
    "unitary_task = sv1.run(OpenQASMProgram(source=program_with_unitary), shots=10)\n",
    "unitary_result = unitary_task.result()\n",
    "print(unitary_result.measurement_counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76f2897ec744dafb",
   "metadata": {},
   "source": [
    "As expected, the Pauli Y gate flipped the initial 0 state to the 1 state."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "111ead44155005a8",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "`from_ir` will reconstruct the unitary:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7b631be4b6785b90",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:54.403015Z",
     "start_time": "2023-11-21T08:32:54.398210Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.+0.j, 0.-1.j],\n",
       "       [0.+1.j, 0.+0.j]])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Circuit.from_ir(program_with_unitary).instructions[0].operator.to_matrix()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce9caf39",
   "metadata": {},
   "source": [
    "## Measuring specific qubits with OpenQASM\n",
    "\n",
    "The Local State Vector Simulator and Local Density Matrix Simulator support submitting OpenQASM programs where a subset of the circuit's qubits can be measured, which is often called <i>partial measurement</i>. For example, you can create a two-qubit circuit and only measure the first qubit like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "d33a0cb6",
   "metadata": {},
   "outputs": [],
   "source": [
    "partial_measure_qasm = \"\"\"\n",
    "OPENQASM 3.0;\n",
    "bit[1] b;\n",
    "qubit[2] q;\n",
    "h q[0];\n",
    "cnot q[0], q[1];\n",
    "b[0] = measure q[0];\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81247da3",
   "metadata": {},
   "source": [
    "We can see that there are two qubits, `q[0]` and `q[1]` but we are only measuring qubit 0 here: `b[0] = measure q[0]`.\n",
    "\n",
    "Now we can run this on the local state vector simulator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d747fcb4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'0': 9, '1': 1})\n",
      "Measured qubits:  [0]\n"
     ]
    }
   ],
   "source": [
    "from braket.devices import LocalSimulator\n",
    "\n",
    "local_sim = LocalSimulator()\n",
    "partial_measure_local_sim_task = local_sim.run(\n",
    "    OpenQASMProgram(source=partial_measure_qasm),\n",
    "    shots=10,\n",
    ")\n",
    "partial_measure_local_sim_result = partial_measure_local_sim_task.result()\n",
    "print(partial_measure_local_sim_result.measurement_counts)\n",
    "print(\"Measured qubits: \", partial_measure_local_sim_result.measured_qubits)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "923c3cd7",
   "metadata": {},
   "source": [
    "As expected, only the targeted qubits were measured. \n",
    "\n",
    "The OpenQASM program `partial_measure_qasm` is equivalent to the following `Circuit`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "5b4e6ae6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'1': 6, '0': 4})\n",
      "Measured qubits:  [0]\n"
     ]
    }
   ],
   "source": [
    "partial_measure = Circuit().h(0).cnot(0, 1).measure(0)\n",
    "partial_measure_task = local_sim.run(partial_measure, shots=10)\n",
    "partial_measure_result = partial_measure_task.result()\n",
    "print(partial_measure_result.measurement_counts)\n",
    "print(\"Measured qubits: \", partial_measure_result.measured_qubits)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49b97636",
   "metadata": {},
   "source": [
    "We can check whether a device supports partial measurement by inspecting the `requiresAllQubitsMeasurement` field in its action properties; if it is `False`, then partial measurement is supported."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "74a715eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "AwsDevice(Devices.Rigetti.Ankaa3).properties.action[\n",
    "    \"braket.ir.openqasm.program\"\n",
    "].requiresAllQubitsMeasurement"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6a94ac1",
   "metadata": {},
   "source": [
    "Here, `requiresAllQubitsMeasurement` is `False`, which indicates that not all qubits must be measured."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8996e0c59741a66f",
   "metadata": {},
   "source": [
    "## Qubit Rewiring with OpenQASM\n",
    "\n",
    "Amazon Braket supports the [physical qubit notation within OpenQASM](https://openqasm.com/language/types.html#physical-qubits) on Rigetti devices. When using physical qubits, you have to ensure that the qubits are indeed connected on the selected device. Alternatively, if qubit registers are used instead, the `PARTIAL` rewiring strategy is enabled by default on Rigetti devices. The following example shows how to use physical qubit notation in an OpenQASM program:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "144cc3e458a21b2d",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:54.410178Z",
     "start_time": "2023-11-21T08:32:54.402877Z"
    }
   },
   "outputs": [],
   "source": [
    "ghz_program_with_physical_qubits = \"\"\"\n",
    "// Prepare a GHZ state\n",
    "OPENQASM 3;\n",
    "\n",
    "bit[3] ro;\n",
    "h $0;\n",
    "cnot $0, $1;\n",
    "cnot $1, $2;\n",
    "ro[0] = measure $0;\n",
    "ro[1] = measure $1;\n",
    "ro[2] = measure $2;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f01e6dc44e0e226",
   "metadata": {},
   "source": [
    "We can run the above program on the Rigetti Ankaa-3 device,"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "768b980b78c4d451",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:55.489511Z",
     "start_time": "2023-11-21T08:32:54.414005Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Measured qubits: [0, 1, 2]\n"
     ]
    }
   ],
   "source": [
    "# choose the quantum device\n",
    "rigetti = AwsDevice(Devices.Rigetti.Ankaa3)\n",
    "\n",
    "ghz_program_with_physical_qubits_task = rigetti.run(\n",
    "    OpenQASMProgram(source=ghz_program_with_physical_qubits),\n",
    "    shots=10,\n",
    ")\n",
    "measured_qubits = ghz_program_with_physical_qubits_task.result().measured_qubits\n",
    "print(\"Measured qubits:\", measured_qubits)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a5597dc7843d541",
   "metadata": {},
   "source": [
    "As we can see, physical qubits 0, 1 and 2 are indeed being used and measured."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25cb2ccc4337dc1e",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-block alert-info\">\n",
    "    <b>Note:</b> This section and the next verbatim box section uses the Rigetti Ankaa-3 device. When you run this notebook, make sure the device is currently available. You can find QPU availability windows on the <a href=\"https://us-west-1.console.aws.amazon.com/braket/home?region=us-west-1#/devices\">Devices page</a> in the Amazon Braket Console\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c30ec68828b17ff",
   "metadata": {},
   "source": [
    "## Verbatim Compilation with OpenQASM\n",
    "\n",
    "In [a previous example notebook](https://github.com/amazon-braket/amazon-braket-examples/blob/main/examples/braket_features/Verbatim_Compilation.ipynb), we talked about verbatim compilation on Braket to gain more precise control on Rigetti devices. With OpenQASM3.0, we can use the `box` syntax together with the `verbatim` pragma to perform verbatim compilation. Here is an example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "a69ab8a7a04aa190",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:55.492060Z",
     "start_time": "2023-11-21T08:32:55.489240Z"
    }
   },
   "outputs": [],
   "source": [
    "program_with_verbatim_box = \"\"\"\n",
    "OPENQASM 3;\n",
    "\n",
    "bit[2] ro;\n",
    "#pragma braket verbatim\n",
    "box{\n",
    " rx(3.141592653589793) $0;\n",
    " rx(3.141592653589793) $0;\n",
    " iswap $0, $1;\n",
    "}\n",
    "ro[0] = measure $0;\n",
    "ro[1] = measure $1;\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d503dc700f00f7b7",
   "metadata": {},
   "source": [
    "To program with verbatim boxes, we need to make sure that\n",
    "- we are using native gates supported by Rigetti devices. Native gates can be found using the following script:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a1cb1cb78d32b7b1",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:55.496864Z",
     "start_time": "2023-11-21T08:32:55.493060Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The native gates for the Ankaa-3 device are:\n",
      "rx\n",
      "rz\n",
      "iswap\n"
     ]
    }
   ],
   "source": [
    "print(\"The native gates for the\", rigetti.name, \"device are:\")\n",
    "for gate in rigetti.properties.paradigm.nativeGateSet:\n",
    "    print(gate)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d24104e29717209",
   "metadata": {},
   "source": [
    "- we use the physical qubit notation.\n",
    "- qubit operands are indeed connected on the physical device. Recall that the device qubit connectivity can be found using the following commands:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "f6a7ffff1914fd03",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:56.296409Z",
     "start_time": "2023-11-21T08:32:55.495523Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'0': ['1', '7'], '1': ['0', '2', '8'], '2': ['1', '3', '9'], '3': ['2', '4', '10'], '4': ['3', '5', '11'], '5': ['4', '6', '12'], '6': ['5', '13'], '7': ['0', '8', '14'], '8': ['1', '7', '9', '15'], '9': ['2', '8', '10', '16'], '10': ['3', '9', '11'], '11': ['4', '10', '12', '18'], '12': ['5', '11', '13', '19'], '13': ['6', '12', '20'], '14': ['7', '15', '21'], '15': ['8', '14', '16', '22'], '16': ['9', '15', '17', '23'], '17': ['16', '18', '24'], '18': ['11', '17', '19', '25'], '19': ['12', '18', '20', '26'], '20': ['13', '19', '27'], '21': ['14', '22', '28'], '22': ['15', '21', '23', '29'], '23': ['16', '22', '24', '30'], '24': ['17', '23', '25', '31'], '25': ['18', '24', '26', '32'], '26': ['19', '25', '27', '33'], '27': ['20', '26'], '28': ['21', '29', '35'], '29': ['22', '28', '30', '36'], '30': ['23', '29', '31', '37'], '31': ['24', '30', '38'], '32': ['25', '33', '39'], '33': ['26', '32', '34', '40'], '34': ['33', '41'], '35': ['28', '36'], '36': ['29', '35', '37', '43'], '37': ['30', '36', '38', '44'], '38': ['31', '37', '39', '45'], '39': ['32', '38', '40', '46'], '40': ['33', '39', '41', '47'], '41': ['34', '40'], '43': ['36', '50'], '44': ['37', '45', '51'], '45': ['38', '44', '46', '52'], '46': ['39', '45', '47', '53'], '47': ['40', '46', '54'], '49': ['50', '56'], '50': ['43', '49', '51', '57'], '51': ['44', '50', '52', '58'], '52': ['45', '51', '53', '59'], '53': ['46', '52', '54', '60'], '54': ['47', '53', '55', '61'], '55': ['54', '62'], '56': ['49', '57', '63'], '57': ['50', '56', '58', '64'], '58': ['51', '57', '59', '65'], '59': ['52', '58', '60', '66'], '60': ['53', '59', '61', '67'], '61': ['54', '60', '62', '68'], '62': ['55', '61', '69'], '63': ['56', '64', '70'], '64': ['57', '63', '65', '71'], '65': ['58', '64', '66', '72'], '66': ['59', '65', '73'], '67': ['60', '68', '74'], '68': ['61', '67', '69', '75'], '69': ['62', '68', '76'], '70': ['63', '71', '77'], '71': ['64', '70', '72', '78'], '72': ['65', '71', '73', '79'], '73': ['66', '72', '74', '80'], '74': ['67', '73', '75', '81'], '75': ['68', '74', '76', '82'], '76': ['69', '75', '83'], '77': ['70', '78'], '78': ['71', '77', '79'], '79': ['72', '78', '80'], '80': ['73', '79', '81'], '81': ['74', '80', '82'], '82': ['75', '81', '83'], '83': ['76', '82']}\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import networkx as nx\n",
    "\n",
    "# access and visualize the device topology\n",
    "print(rigetti.properties.paradigm.connectivity.connectivityGraph)\n",
    "nx.draw_kamada_kawai(rigetti.topology_graph, with_labels=True, font_color=\"white\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fa09f83a0747394",
   "metadata": {},
   "source": [
    "Now we can submit a quantum task of the above program with verbatim box."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "b5e7c89c1c9f8879",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:56.299544Z",
     "start_time": "2023-11-21T08:32:56.297194Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PRAGMA INITIAL_REWIRING \"NAIVE\"\n",
      "DECLARE ro BIT[2]\n",
      "PRAGMA PRESERVE_BLOCK\n",
      "RX(3.141592653589793) 0\n",
      "RX(3.141592653589793) 0\n",
      "ISWAP 0 1\n",
      "PRAGMA END_PRESERVE_BLOCK\n",
      "MEASURE 0 ro[0]\n",
      "MEASURE 1 ro[1]\n"
     ]
    }
   ],
   "source": [
    "verbatim_task = rigetti.run(OpenQASMProgram(source=program_with_verbatim_box), shots=10)\n",
    "verbatim_result = verbatim_task.result()\n",
    "meta = verbatim_result.additional_metadata.rigettiMetadata\n",
    "print(meta.compiledProgram)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bcd27bc239db889d",
   "metadata": {},
   "source": [
    "As shown above, the two consecutive `rx`  $\\pi$-rotation gates did not get optimized and we confirm that our program was indeed executed verbatim."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4768a96f7b1b3b6",
   "metadata": {},
   "source": [
    "## Requesting Result Types with OpenQASM\n",
    "\n",
    "Braket provides [a rich library of result types](https://docs.aws.amazon.com/braket/latest/developerguide/braket-result-types.html) for circuit executions. With OpenQASM, requesting different result types for our quantum tasks is easier than ever using the `result` pragma. Next, we give an example of requesting result types for our Bell state program submitted to SV1. Before doing that, let's see what result types are supported on SV1:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b17d21060cda7fa7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:56.304110Z",
     "start_time": "2023-11-21T08:32:56.300283Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "name='Sample' observables=['x', 'y', 'z', 'h', 'i', 'hermitian'] minShots=1 maxShots=100000\n",
      "name='Expectation' observables=['x', 'y', 'z', 'h', 'i', 'hermitian'] minShots=0 maxShots=100000\n",
      "name='Variance' observables=['x', 'y', 'z', 'h', 'i', 'hermitian'] minShots=0 maxShots=100000\n",
      "name='Probability' observables=None minShots=1 maxShots=100000\n",
      "name='Amplitude' observables=None minShots=0 maxShots=0\n",
      "name='AdjointGradient' observables=['x', 'y', 'z', 'h', 'i'] minShots=0 maxShots=0\n"
     ]
    }
   ],
   "source": [
    "# print the result types supported by SV1\n",
    "for iter in sv1.properties.action[\"braket.ir.openqasm.program\"].supportedResultTypes:\n",
    "    print(iter)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bbc02495939691c9",
   "metadata": {},
   "source": [
    "With knowing the supported result types on SV1, we choose to request the `Expectation` of $X \\otimes Z$ observable on `q[0]` and `q[1]` and the `Amplitude` result type for a `shots=0` quantum task of our bell program:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "6e3b521b7356b4bd",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:56.305634Z",
     "start_time": "2023-11-21T08:32:56.302835Z"
    }
   },
   "outputs": [],
   "source": [
    "bell_with_result_type = \"\"\"\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit[2] q;\n",
    "\n",
    "#pragma braket result expectation x(q[0]) @ z(q[1])\n",
    "#pragma braket result amplitude \"00\", \"11\"\n",
    "h q[0];\n",
    "cnot q[0], q[1];\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6435b34ab2525039",
   "metadata": {},
   "source": [
    "The location of the `result` pragma is very flexible as long as it's after the qubit register definition (if you use physical qubits, you can put `result` pragmas anywhere after the program header).\n",
    "\n",
    "We can submit the above program and receive the results for our requested result types."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "a42b0ebc71b59136",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:59.113501Z",
     "start_time": "2023-11-21T08:32:56.306452Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0, {'00': (0.7071067811865475+0j), '11': (0.7071067811865475+0j)}]\n"
     ]
    }
   ],
   "source": [
    "bell_result_types_task = sv1.run(OpenQASMProgram(source=bell_with_result_type), shots=0)\n",
    "bell_result = bell_result_types_task.result()\n",
    "values = bell_result.values\n",
    "print(values)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38675f5678613895",
   "metadata": {},
   "source": [
    "At last, we want to remind our Braket OpenQASM users that there are two requirements when requesting result types:\n",
    "1. For `shots=0` quantum tasks, requesting non-simultaneously measurable result types is allowed, but for `shots>0` quantum tasks, it is not allowed. For example, we can write the following OpenQASM program in a `shots=0` quantum task but not in a `shots>0` quantum task, since the two result types are not simultaneously measurable:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "45277f89209faec7",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:59.115339Z",
     "start_time": "2023-11-21T08:32:59.112696Z"
    }
   },
   "outputs": [],
   "source": [
    "program_with_non_simultaneously_measurable_result_types = \"\"\"\n",
    "OPENQASM 3;\n",
    "\n",
    "qubit[2] q;\n",
    "\n",
    "h q[0];\n",
    "cnot q[0], q[1];\n",
    "\n",
    "#pragma braket result expectation x(q[0]) @ z(q[1])\n",
    "#pragma braket result expectation hermitian([[0, -1im], [1im, 0]]) q[0]\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9af8819fba31a90",
   "metadata": {},
   "source": [
    "2. Do not use measurement instructions and request result types in the same OpenQASM program, otherwise a validation error will be raised. Since measurement instructions are basically equivalent to `#pragma braket result sample z(qubit)`, we encourage users to adapt a consistent style of requesting result types in the same program."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db79ad324a5f6478",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "Circuits constructed with `from_ir` will have the correct result types:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "f3c438da3a351153",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:59.127212Z",
     "start_time": "2023-11-21T08:32:59.124909Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bell_with_result_type:\n",
      "T  : │  0  │  1  │    Result Types    │\n",
      "      ┌───┐       ┌──────────────────┐ \n",
      "q0 : ─┤ H ├───●───┤ Expectation(X@Z) ├─\n",
      "      └───┘   │   └────────┬─────────┘ \n",
      "            ┌─┴─┐ ┌────────┴─────────┐ \n",
      "q1 : ───────┤ X ├─┤ Expectation(X@Z) ├─\n",
      "            └───┘ └──────────────────┘ \n",
      "T  : │  0  │  1  │    Result Types    │\n",
      "\n",
      "Additional result types: Amplitude(00,11)\n"
     ]
    }
   ],
   "source": [
    "print(\"bell_with_result_type:\")\n",
    "print(Circuit.from_ir(bell_with_result_type))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c371de3097c0598a",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "## Advanced OpenQASM features\n",
    "\n",
    "OpenQASM has features beyond what is natively supported by the `Circuit` class. You can run OpenQASM tasks directly on the Local Simulator or use Circuit.from_ir to convert an OpenQASM program to a Circuit object, which is supported by all circuit-based Braket devices. The following OpenQASM program is a GHZ state written with classical control flow and subroutines:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "5acb72003cd0a0f5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:04:08.355878Z",
     "start_time": "2023-12-04T04:04:08.353679Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "ghz_with_advanced_features = \"\"\"\n",
    "OPENQASM 3.0;\n",
    "\n",
    "def ghz(int[32] n) {\n",
    "    h q[0];\n",
    "    for int i in [0:n - 1] {\n",
    "        cnot q[i], q[i + 1];\n",
    "    }\n",
    "}\n",
    "\n",
    "const int[32] n = 5;\n",
    "bit[n + 1] c;\n",
    "qubit[n + 1] q;\n",
    "\n",
    "ghz(n);\n",
    "\n",
    "c = measure q;\n",
    "\"\"\"\n",
    "qasm_program = OpenQASMProgram(source=ghz_with_advanced_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f5d32bd10c64543",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "The local simulator supports these features:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "c82bee530da348f6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:04:13.932270Z",
     "start_time": "2023-12-04T04:04:13.863203Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "This program uses OpenQASM language features that may not be supported on QPUs or on-demand simulators.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Counter({'111111': 503, '000000': 497})"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from braket.devices import LocalSimulator\n",
    "\n",
    "LocalSimulator().run(qasm_program, shots=1000).result().measurement_counts"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96fca192b4f51797",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "but SV1 does not. However, `from_ir` will \"unroll\" the subroutine and loop to create a `Circuit` object that _can_ run on SV1:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "b0886dfc74978bfa",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-12-04T04:04:19.974386Z",
     "start_time": "2023-12-04T04:04:16.705969Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "This program uses OpenQASM language features that may not be supported on QPUs or on-demand simulators.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "T  : │  0  │  1  │  2  │  3  │  4  │  5  │  6  │\n",
      "      ┌───┐                               ┌───┐ \n",
      "q0 : ─┤ H ├───●───────────────────────────┤ M ├─\n",
      "      └───┘   │                           └───┘ \n",
      "            ┌─┴─┐                         ┌───┐ \n",
      "q1 : ───────┤ X ├───●─────────────────────┤ M ├─\n",
      "            └───┘   │                     └───┘ \n",
      "                  ┌─┴─┐                   ┌───┐ \n",
      "q2 : ─────────────┤ X ├───●───────────────┤ M ├─\n",
      "                  └───┘   │               └───┘ \n",
      "                        ┌─┴─┐             ┌───┐ \n",
      "q3 : ───────────────────┤ X ├───●─────────┤ M ├─\n",
      "                        └───┘   │         └───┘ \n",
      "                              ┌─┴─┐       ┌───┐ \n",
      "q4 : ─────────────────────────┤ X ├───●───┤ M ├─\n",
      "                              └───┘   │   └───┘ \n",
      "                                    ┌─┴─┐ ┌───┐ \n",
      "q5 : ───────────────────────────────┤ X ├─┤ M ├─\n",
      "                                    └───┘ └───┘ \n",
      "T  : │  0  │  1  │  2  │  3  │  4  │  5  │  6  │\n",
      "Counter({'111111': 518, '000000': 482})\n"
     ]
    }
   ],
   "source": [
    "circuit = Circuit.from_ir(qasm_program)\n",
    "print(circuit)\n",
    "print(sv1.run(circuit, shots=1000).result().measurement_counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a329fa3df47ffee4",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "For an in-depth exploration of advanced OpenQASM features, see [Simulating Advanced OpenQASM Programs with the Local Simulator](https://github.com/amazon-braket/amazon-braket-examples/blob/main/examples/braket_features/Simulating_Advanced_OpenQASM_Programs_with_the_Local_Simulator.ipynb)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e4c7a48a2b83ce1",
   "metadata": {},
   "source": [
    "# Conclusion\n",
    "\n",
    "In this notebook, you learned how to submit OpenQASM quantum tasks and use OpenQASM features on Braket. Hope you enjoyed it! You can find more information about OpenQASM3.0 in its [live specification](https://openqasm.com/), and you can learn more about OpenQASM support on Braket in the [Amazon Braket documentation](https://docs.aws.amazon.com/braket/latest/developerguide/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "3f5bc3dfad520f14",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2023-11-21T08:32:59.624204Z",
     "start_time": "2023-11-21T08:32:59.127772Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Quantum Task Summary\n",
      "{<_Amazon.SV1: 'arn:aws:braket:::device/quantum-simulator/amazon/sv1'>: {'shots': 1120, 'tasks': {'CREATED': 1, 'COMPLETED': 4}, 'execution_duration': datetime.timedelta(microseconds=11000), 'billed_execution_duration': datetime.timedelta(seconds=12)}, <_Amazon.DM1: 'arn:aws:braket:::device/quantum-simulator/amazon/dm1'>: {'shots': 30, 'tasks': {'COMPLETED': 3}, 'execution_duration': datetime.timedelta(microseconds=6000), 'billed_execution_duration': datetime.timedelta(seconds=9)}, <_Rigetti.Ankaa3: 'arn:aws:braket:us-west-1::device/qpu/rigetti/Ankaa-3'>: {'shots': 20, 'tasks': {'COMPLETED': 2}}}\n",
      "Note: Charges shown are estimates based on your Amazon Braket simulator and quantum processing unit (QPU) task usage. Estimated charges shown may differ from your actual charges. Estimated charges do not factor in any discounts or credits, and you may experience additional charges based on your use of other services such as Amazon Elastic Compute Cloud (Amazon EC2).\n",
      "Estimated cost to run this example: 0.64 USD\n"
     ]
    }
   ],
   "source": [
    "print(\"Quantum Task Summary\")\n",
    "print(t.quantum_tasks_statistics())\n",
    "print(\n",
    "    \"Note: Charges shown are estimates based on your Amazon Braket simulator and quantum processing unit (QPU) task usage. Estimated charges shown may differ from your actual charges. Estimated charges do not factor in any discounts or credits, and you may experience additional charges based on your use of other services such as Amazon Elastic Compute Cloud (Amazon EC2).\",\n",
    ")\n",
    "print(\n",
    "    f\"Estimated cost to run this example: {t.qpu_tasks_cost() + t.simulator_tasks_cost():.2f} USD\",\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "3.10.10",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}