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",
    "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",
    "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/Aspen-M-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": 8,
   "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|\n",
      "          \n",
      "q0 : -H-C-\n",
      "        | \n",
      "q1 : ---X-\n",
      "\n",
      "T  : |0|1|\n"
     ]
    }
   ],
   "source": [
    "from braket.circuits import Circuit\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": 9,
   "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": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# supportedOperations\n",
    "sv1.properties.action['braket.ir.openqasm.program'].supportedOperations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# supportedPragmas\n",
    "sv1.properties.action['braket.ir.openqasm.program'].supportedPragmas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": 11,
     "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": 12,
   "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": 13,
   "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": 13,
     "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": 14,
   "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": 15,
   "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({'111': 5, '000': 4, '100': 1})\n",
      "openqasm measurement results: Counter({'000': 5, '111': 4, '100': 1})\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": 16,
   "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": 17,
   "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     |\n",
      "                                                    \n",
      "q0 : -H-DEPO(0.1)-C-DEPO(0.1)-DEPO(0.1)-------------\n",
      "                  |                                 \n",
      "q1 : -------------X-DEPO(0.1)-----------C-DEPO(0.1)-\n",
      "                                        |           \n",
      "q2 : -----------------------------------X-----------\n",
      "\n",
      "T  : |     0     |          1          |     2     |\n",
      "\n",
      "noisy_program_with_kraus_operators:\n",
      "T  : | 0  |1|\n",
      "             \n",
      "q0 : -H-KR-C-\n",
      "           | \n",
      "q1 : ------X-\n",
      "\n",
      "T  : | 0  |1|\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": 18,
   "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": 19,
   "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({'00': 4, '01': 3, '11': 3})\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": 20,
   "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": 21,
   "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({'00': 3, '01': 3, '11': 2, '10': 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": 22,
   "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": 23,
   "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": 24,
   "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": 24,
     "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": 25,
   "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": 27,
   "id": "d747fcb4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'0': 7, '1': 3})\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(OpenQASMProgram(source=partial_measure_qasm), shots = 10)\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": 43,
   "id": "5b4e6ae6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Counter({'0': 7, '1': 3})\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": 15,
   "id": "74a715eb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "AwsDevice(Devices.Rigetti.AspenM3).properties.action['braket.ir.openqasm.program'].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": 24,
   "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 Aspen-M-3 device,"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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",
    "aspen_m = AwsDevice(Devices.Rigetti.AspenM3)\n",
    "\n",
    "ghz_program_with_physical_qubits_task = aspen_m.run(OpenQASMProgram(source=ghz_program_with_physical_qubits), shots = 10)\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 Aspen-M-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": 26,
   "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",
    " cz $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": 27,
   "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 Aspen-M-3 device are:\n",
      "rx\n",
      "rz\n",
      "cz\n",
      "cphaseshift\n",
      "xy\n"
     ]
    }
   ],
   "source": [
    "print(\"The native gates for the\", aspen_m.name, \"device are:\")\n",
    "for gate in aspen_m.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": 28,
   "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', '103'], '100': ['101', '107'], '10': ['11', '17', '113'], '110': ['111', '117'], '20': ['21', '27', '123'], '120': ['121', '127'], '30': ['31', '37', '133'], '130': ['131', '137'], '40': ['41', '47', '143'], '140': ['141', '147'], '1': ['0', '2', '16'], '101': ['100', '102'], '11': ['10', '12'], '111': ['110', '112', '126'], '21': ['20', '22', '36'], '121': ['120', '122'], '31': ['30', '32', '46'], '131': ['130', '132', '146'], '41': ['40', '42'], '141': ['140', '142'], '2': ['1', '3', '15'], '102': ['101', '103', '115'], '12': ['11', '13', '25'], '112': ['111', '113', '125'], '22': ['21', '23', '35'], '122': ['121', '123', '135'], '32': ['31', '33', '45'], '132': ['131', '133', '145'], '42': ['41', '43'], '142': ['141', '143'], '3': ['2', '4'], '103': ['102', '104', '0'], '13': ['12', '14'], '113': ['112', '114', '10'], '23': ['22', '24'], '123': ['122', '124', '20'], '33': ['32', '34'], '133': ['132', '134', '30'], '43': ['42', '44'], '143': ['142', '144', '40'], '4': ['3', '5'], '104': ['103', '105', '7'], '14': ['13', '15'], '114': ['113', '115', '17'], '24': ['23', '25'], '124': ['123', '125', '27'], '34': ['33', '35'], '134': ['133', '135', '37'], '44': ['43', '45'], '144': ['143', '145', '47'], '5': ['4', '6'], '105': ['104', '106'], '15': ['14', '16', '2'], '115': ['114', '102'], '25': ['24', '26', '12'], '125': ['124', '126', '112'], '35': ['34', '36', '22'], '135': ['134', '122'], '45': ['44', '46', '32'], '145': ['144', '146', '132'], '6': ['5'], '106': ['105', '107'], '16': ['15', '17', '1'], '116': ['117'], '26': ['25', '27'], '126': ['125', '127', '111'], '36': ['35', '37', '21'], '46': ['45', '47', '31'], '146': ['145', '147', '131'], '7': ['0', '104'], '107': ['106', '100'], '17': ['16', '10', '114'], '117': ['116', '110'], '27': ['26', '20', '124'], '127': ['126', '120'], '37': ['36', '30', '134'], '137': ['130'], '47': ['46', '40', '144'], '147': ['146', '140']}\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",
    "# access and visualize the device topology\n",
    "print(aspen_m.properties.paradigm.connectivity.connectivityGraph)\n",
    "nx.draw_kamada_kawai(aspen_m.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": 29,
   "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": [
      "\n",
      "DECLARE ro BIT[2]\n",
      "PRAGMA INITIAL_REWIRING \"NAIVE\"\n",
      "RESET\n",
      "RX(pi) 0\n",
      "RX(pi) 0\n",
      "CZ 0 1\n",
      "MEASURE 1 ro[1]\n",
      "MEASURE 0 ro[0]\n"
     ]
    }
   ],
   "source": [
    "verbatim_task = aspen_m.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": 30,
   "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": 31,
   "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": 32,
   "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": 33,
   "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": 34,
   "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-C-Expectation(X@Z)-\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": 35,
   "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",
    "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": 36,
   "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({'000000': 480, '111111': 520})"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from braket.devices import LocalSimulator\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": 37,
   "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|\n",
      "                  \n",
      "q0 : -H-C---------\n",
      "        |         \n",
      "q1 : ---X-C-------\n",
      "          |       \n",
      "q2 : -----X-C-----\n",
      "            |     \n",
      "q3 : -------X-C---\n",
      "              |   \n",
      "q4 : ---------X-C-\n",
      "                | \n",
      "q5 : -----------X-\n",
      "\n",
      "T  : |0|1|2|3|4|5|\n",
      "Counter({'000000': 503, '111111': 497})\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": 35,
   "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",
      "{'arn:aws:braket:::device/quantum-simulator/amazon/sv1': {'shots': 120, 'tasks': {'CREATED': 1, 'COMPLETED': 3}, 'execution_duration': datetime.timedelta(microseconds=20000), 'billed_execution_duration': datetime.timedelta(seconds=9)}, 'arn:aws:braket:::device/quantum-simulator/amazon/dm1': {'shots': 30, 'tasks': {'COMPLETED': 3}, 'execution_duration': datetime.timedelta(microseconds=12000), 'billed_execution_duration': datetime.timedelta(seconds=9)}}\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: 0.02 USD\n"
     ]
    }
   ],
   "source": [
    "print(\"Quantum Task Summary\")\n",
    "print(t.quantum_tasks_statistics())\n",
    "print('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",
    "print(f\"Estimated cost to run this example: {t.qpu_tasks_cost() + t.simulator_tasks_cost():.2f} USD\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}