Option to load custom RE parameters in DF chemistry sample (#1662)

This adds a program argument `-p` (long `--paramsfile`) to the Python
script of the DF chemistry sample to have the option to load one or
multiple RE configurations that are stored as objects or arrays in JSON
files. If no such argument is specified, the default configuration is
used as before.

This addresses the second bullet point in #1115.

samples/estimation/df-chemistry/README.md

@@ -1,10 +1,10 @@
 # Resource Estimation for Double-factorized Chemistry
 
-In this sample we evaluate the physical resource estimates of using the so-called double-factorized qubitization algorithm described in [[Phys. Rev. Research 3, 033055 (2021)](https://doi.org/10.1103/PhysRevResearch.3.033055)] to calculate the energy of a user provided Hamiltonian to chemical accuracy of 1 mHa. 
+In this sample we evaluate the physical resource estimates of using the so-called double-factorized qubitization algorithm described in [[Phys. Rev. Research 3, 033055 (2021)](https://doi.org/10.1103/PhysRevResearch.3.033055)] to calculate the energy of a user provided Hamiltonian to chemical accuracy of 1 mHa.
 
 The Hamiltonian is provided as an FCIDUMP file that is available on your machine or can be downloaded via an HTTPS URL.
 
-```
+```text
 usage: chemistry.py [-h] [-f FCIDUMPFILE]
 
 Double-factorized chemistry sample
@@ -13,20 +13,24 @@ options:
   -h, --help            show this help message and exit
   -f FCIDUMPFILE, --fcidumpfile FCIDUMPFILE
                         Path to the FCIDUMP file describing the Hamiltonian
+  -p [PARAMSFILE ...], --paramsfile [PARAMSFILE ...]
+                        Optional parameter files to use for estimation
 ```
 
 For example, the following command will download the FCIDUMP file `n2-10e-8o` to the working folder and run resource estimation for it:
 
-```
+```shell
 chemistry.py -f https://aka.ms/fcidump/n2-10e-8o
 ```
 
 After that, you can pass the path to the downloaded file to the script instead:
 
-```
+```shell
 chemistry.py -f n2-10e-8o
 ```
 
+By default, physical resources are estimates for a Majorana based qubit with 10⁻⁶ error rates (`qubit_maj_ns_e6`), a Floquet code QEC scheme.  The error budget is set to 0.01 to reach the required chemical accuracy of 1 mHa.  The `-p` program argument can be used to load other resource estimation parameters, specified in JSON files.  The JSON files can either contain a JSON object for one configuration, or an array of JSON objects for multiple configurations.  Make sure to set the error budget to 0.01 to guarantee the correct chemical accuracy.
+
 You can choose some of the following URLs to download example files:
 
 | URL  | Instance name  | Description  |
@@ -37,4 +41,4 @@ You can choose some of the following URLs to download example files:
 | https://aka.ms/fcidump/fe2s2-10e-40o         | fe2s2-10e-40o         | 10 electron, 40 orbital active space of [2Fe, 2S] cluster that is shown in [this paper](https://www.nature.com/articles/nchem.2041) |
 | https://aka.ms/fcidump/polyyne-24e-24o       | polyyne-24e-24o       | 24 electron, 24 orbital active space of the polyyne molecule |
 
-The numbers for the XVIII-cas4-fb-64e-56o instance roughly match the numbers in the paper [Assessing requirements for scaling quantum computers to real-world impact](https://aka.ms/AQ/RE/Paper), as we incorporated a few improvements in the implementation of the double-factorized chemistry algorithm as compared to the version used when the paper was published.
+The numbers for the XVIII-cas4-fb-64e-56o instance roughly match the numbers in the paper [Assessing requirements for scaling quantum computers to real-world impact](https://aka.ms/AQ/RE/Paper), as we incorporated a few improvements in the implementation of the double-factorized chemistry algorithm as compared to the version used when the paper was published.

samples/estimation/df-chemistry/chemistry.py

@@ -1,5 +1,6 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
+import json
 import math
 import numpy as np
 import numpy.typing as npt
@@ -445,6 +446,12 @@ def ndarray2d_to_string(arr):
     default="https://aka.ms/fcidump/n2-10e-8o",
     help="Path to the FCIDUMP file describing the Hamiltonian",
 )
+parser.add_argument(
+    "-p",
+    "--paramsfile",
+    nargs="*",
+    help="Optional parameter files to use for estimation",
+)
 args = parser.parse_args()
 
 # ----- Read the FCIDUMP file and get resource estimates from Q# algorithm -----
@@ -479,23 +486,37 @@ def ndarray2d_to_string(arr):
     "Microsoft.Quantum.Applications.Chemistry.DoubleFactorizedChemistryParameters(0.001,))"
 )
 
-# Get resource estimates
-res = qsharp.estimate(
-    qsharp_string,
-    params={
+# Collect resource estimation parameters
+if args.paramsfile is None:
+    params = {
         "errorBudget": 0.01,
         "qubitParams": {"name": "qubit_maj_ns_e6"},
         "qecScheme": {"name": "floquet_code"},
-    },
+    }
+else:
+    params = []
+    for paramsfile in args.paramsfile:
+        with open(paramsfile) as f:
+            data = json.load(f)
+            if isinstance(data, dict):
+                params.append(data)
+            else:
+                params += data
+
+# Get resource estimates
+res = qsharp.estimate(
+    qsharp_string,
+    params=params,
 )
 
 # Store estimates in json file
 with open("resource_estimate.json", "w") as f:
     f.write(res.json)
 
 # Print high-level resource estimation results
-print(f"Algorithm runtime: {res['physicalCountsFormatted']['runtime']}")
-print(
-    f"Number of physical qubits required: {res['physicalCountsFormatted']['physicalQubits']}"
-)
+if "physicalCountsFormatted" in res:
+    print(f"Algorithm runtime: {res['physicalCountsFormatted']['runtime']}")
+    print(
+        f"Number of physical qubits required: {res['physicalCountsFormatted']['physicalQubits']}"
+    )
 print("For more detailed resource counts, see file resource_estimate.json")