Minor fixes for markdown code examples (Qiskit#666)

Minor tweaks to make the code examples run. Closes Qiskit#609.

docs/build/interoperate-qiskit-qasm2.mdx

@@ -13,6 +13,7 @@ Qiskit provides some tools for converting between OpenQASM representations of qu
 Currently two high-level functions are available for importing from OpenQASM 2 into Qiskit. These functions are `qasm2.load()`, which takes a file name, and `qasm2.loads()`, which takes the program itself as a string. 
 
 ```python
+import qiskit.qasm2
 qiskit.qasm2.load(filename, *, include_path=('.',), include_input_directory='append', custom_instructions=(), custom_classical=(), strict=False)
 ```
 
@@ -110,7 +111,7 @@ Use `qasm2.loads()` to import an OpenQASM 2 program as a string into a QuantumCi
 
 ```python
 import math
-from qiskit.qasm2 
+import qiskit.qasm2
 
 program = '''
     include "qelib1.inc";

docs/build/interoperate-qiskit-qasm3.mdx

@@ -14,22 +14,25 @@ This function is still in the exploratory phase.  Therefore, it is likely that t
 
 ## Import an OpenQASM 3 program into Qiskit
 
+You must install the package `qiskit_qasm3_import ` to use this function. Install using the following command.
+
+```python
+pip install qiskit-qasm3-import
+```
+
 Currently two high-level functions are available for importing from OpenQASM 3 into Qiskit. These functions are `load()`, which takes a file name, and `loads()`, which takes the program itself as a string: 
 
 
 ```python
+import qiskit.qasm3
 qiskit.qasm3.load(file_name)
 ```
 
 ```python
+import qiskit.qasm3
 qiskit.qasm3.loads(program-string)
 ```
 
-You must install the package `qiskit_qasm3_import ` to use this function.  For example:  
-
-```python
-pip install qiskit-qasm3-import
-```
 
 In this example, we define a quantum program using OpenQASM 3, and use `loads()` to directly convert it into a QuantumCircuit:
 

docs/transpile/common-parameters.mdx

@@ -19,13 +19,17 @@ When the approximation degree is less than 1.0, circuits with one or two CX gate
 As an example, we generate a random 2-qubit `UnitaryGate` which will be synthesized in the initial stage. Setting the `approximation_degree` less than 1.0 might generate an approximate circuit. We must also specify the `basis_gates` to let the synthesis method know which gates it can use for the approximate synthesis.
 
 ```python
-    UU = random_unitary(4, seed=12345)
-    rand_U = UnitaryGate(UU)
+from qiskit import QuantumCircuit, transpile
+from qiskit.circuit.library import UnitaryGate
+from qiskit.quantum_info import random_unitary
 
-    qc = QuantumCircuit(2)
-    qc.append(rand_U, range(2))
-    approx_qc = transpile(qc, approximation_degree=0.85, basis_gates=["sx", "rz", "cx"])
-    print(approx_qc.count_ops()["cx"])
+UU = random_unitary(4, seed=12345)
+rand_U = UnitaryGate(UU)
+
+qc = QuantumCircuit(2)
+qc.append(rand_U, range(2))
+approx_qc = transpile(qc, approximation_degree=0.85, basis_gates=["sx", "rz", "cx"])
+print(approx_qc.count_ops()["cx"])
 ```
 
 This yields an output of `2` because the approximation requires fewer CX gates.
@@ -38,7 +42,7 @@ The seed transpiler argument sets the random seed for the stochastic parts of th
 Example:
 
 ```python
-    optimized_1 = transpile(qc, backend=backend, seed_transpiler=11, optimization_level=1)
+optimized_1 = transpile(qc, seed_transpiler=11, optimization_level=1)
 ```
 
 <span id="init-layout"></span>