Merge pull request pybamm-team#4441 from pybamm-team/telemetry

Add telemetry

.gitignore

@@ -46,7 +46,6 @@ input/*
 
 # simulation outputs
 out/
-config.py
 matplotlibrc
 *.pickle
 *.sav

.lycheeignore

@@ -15,3 +15,6 @@ file:///home/runner/work/PyBaMM/PyBaMM/docs/source/user_guide/fundamentals/pybam
 
 # Errors in docs/source/user_guide/index.md
 file:///home/runner/work/PyBaMM/PyBaMM/docs/source/user_guide/api_docs
+
+# Telemetry
+https://us.i.posthog.com

CHANGELOG.md

@@ -6,11 +6,11 @@
 - Adds an option "voltage as a state" that can be "false" (default) or "true". If "true" adds an explicit algebraic equation for the voltage. ([#4507](https://github.com/pybamm-team/PyBaMM/pull/4507))
 - Improved `QuickPlot` accuracy for simulations with Hermite interpolation. ([#4483](https://github.com/pybamm-team/PyBaMM/pull/4483))
 - Added Hermite interpolation to the (`IDAKLUSolver`) that improves the accuracy and performance of post-processing variables. ([#4464](https://github.com/pybamm-team/PyBaMM/pull/4464))
+- Added basic telemetry to record which functions are being run. See [Telemetry section in the User Guide](https://docs.pybamm.org/en/latest/source/user_guide/index.html#telemetry) for more information. ([#4441](https://github.com/pybamm-team/PyBaMM/pull/4441))
 - Added `BasicDFN` model for sodium-ion batteries ([#4451](https://github.com/pybamm-team/PyBaMM/pull/4451))
 - Added sensitivity calculation support for `pybamm.Simulation` and `pybamm.Experiment` ([#4415](https://github.com/pybamm-team/PyBaMM/pull/4415))
 - Added OpenMP parallelization to IDAKLU solver for lists of input parameters ([#4449](https://github.com/pybamm-team/PyBaMM/pull/4449))
-- Added phase-dependent particle options to LAM
-  ([#4369](https://github.com/pybamm-team/PyBaMM/pull/4369))
+- Added phase-dependent particle options to LAM ([#4369](https://github.com/pybamm-team/PyBaMM/pull/4369))
 - Added a lithium ion equivalent circuit model with split open circuit voltages for each electrode (`SplitOCVR`). ([#4330](https://github.com/pybamm-team/PyBaMM/pull/4330))
 - Added the `pybamm.DiscreteTimeSum` expression node to sum an expression over a sequence of data times, and accompanying `pybamm.DiscreteTimeData` class to store the data times and values ([#4501](https://github.com/pybamm-team/PyBaMM/pull/4501))
 

conftest.py

@@ -51,3 +51,8 @@ def set_random_seed():
 @pytest.fixture(autouse=True)
 def set_debug_value():
     pybamm.settings.debug_mode = True
+
+
+@pytest.fixture(autouse=True)
+def disable_telemetry():
+    pybamm.telemetry.disable()

docs/source/user_guide/index.md

@@ -73,3 +73,12 @@ glob:
 ../examples/notebooks/creating_models/5-half-cell-model.ipynb
 ../examples/notebooks/creating_models/6-a-simple-SEI-model.ipynb
 ```
+
+# Telemetry
+
+PyBaMM optionally collects anonymous usage data to help improve the library. This telemetry is opt-in and can be easily disabled. Here's what you need to know:
+
+- **What is collected**: Basic usage information like PyBaMM version, Python version, and which functions are run.
+- **Why**: To understand how PyBaMM is used and prioritize development efforts.
+- **Opt-out**: To disable telemetry, set the environment variable `PYBAMM_DISABLE_TELEMETRY=true` (or any value other than `false`) or use `pybamm.telemetry.disable()` in your code.
+- **Privacy**: No personal information (name, email, etc) or sensitive information (parameter values, simulation results, etc) is ever collected.

docs/source/user_guide/installation/index.rst

@@ -71,6 +71,7 @@ Package                                                             Minimum supp
 `typing-extensions <https://pypi.org/project/typing-extensions/>`__ 4.10.0
 `pandas <https://pypi.org/project/pandas/>`__                       1.5.0
 `pooch <https://www.fatiando.org/pooch/>`__                         1.8.1
+`posthog <https://posthog.com/>`__                                  3.6.5
 =================================================================== ==========================
 
 .. _install.optional_dependencies:

noxfile.py

@@ -49,6 +49,7 @@ def set_iree_state():
     "IREE_INDEX_URL": os.getenv(
         "IREE_INDEX_URL", "https://iree.dev/pip-release-links.html"
     ),
+    "PYBAMM_DISABLE_TELEMETRY": "true",
 }
 VENV_DIR = Path("./venv").resolve()
 

pyproject.toml

@@ -44,6 +44,7 @@ dependencies = [
     "typing-extensions>=4.10.0",
     "pandas>=1.5.0",
     "pooch>=1.8.1",
+    "posthog",
 ]
 
 [project.urls]

src/pybamm/__init__.py

@@ -194,8 +194,8 @@
 # Batch Study
 from .batch_study import BatchStudy
 
-# Callbacks
-from . import callbacks
+# Callbacks, telemetry, config
+from . import callbacks, telemetry, config
 
 # Pybamm Data manager using pooch
 from .pybamm_data import DataLoader
@@ -204,12 +204,14 @@
 import os
 import pathlib
 import sysconfig
-os.environ["CASADIPATH"] = str(pathlib.Path(sysconfig.get_path('purelib')) / 'casadi')
+
+os.environ["CASADIPATH"] = str(pathlib.Path(sysconfig.get_path("purelib")) / "casadi")
 
 __all__ = [
     "batch_study",
     "callbacks",
     "citations",
+    "config",
     "discretisations",
     "doc_utils",
     "experiment",
@@ -225,8 +227,11 @@
     "simulation",
     "solvers",
     "spatial_methods",
+    "telemetry",
     "type_definitions",
     "util",
     "version",
     "pybamm_data",
 ]
+
+pybamm.config.generate()

src/pybamm/config.py

@@ -0,0 +1,163 @@
+import uuid
+import os
+import platformdirs
+from pathlib import Path
+import pybamm
+import sys
+import threading
+import time
+
+
+def is_running_tests():  # pragma: no cover
+    """
+    Detect if the code is being run as part of a test suite or building docs with Sphinx.
+
+    Returns:
+        bool: True if running tests or building docs, False otherwise.
+    """
+    # Check if pytest or unittest is running
+    if any(
+        test_module in sys.modules for test_module in ["pytest", "unittest", "nose"]
+    ):
+        return True
+
+    # Check for GitHub Actions environment variable
+    if "GITHUB_ACTIONS" in os.environ:
+        return True
+
+    # Check for other common CI environment variables
+    ci_env_vars = ["CI", "TRAVIS", "CIRCLECI", "JENKINS_URL", "GITLAB_CI"]
+    if any(var in os.environ for var in ci_env_vars):
+        return True
+
+    # Check for common test runner names in command-line arguments
+    test_runners = ["pytest", "unittest", "nose", "trial", "nox", "tox"]
+    if any(runner in arg.lower() for arg in sys.argv for runner in test_runners):
+        return True
+
+    # Check if building docs with Sphinx
+    if any(mod == "sphinx" or mod.startswith("sphinx.") for mod in sys.modules):
+        print(
+            f"Found Sphinx module: {[mod for mod in sys.modules if mod.startswith('sphinx')]}"
+        )
+        return True
+
+    return False
+
+
+def ask_user_opt_in(timeout=10):
+    """
+    Ask the user if they want to opt in to telemetry.
+
+    Parameters
+    ----------
+    timeout : float, optional
+        The timeout for the user to respond to the prompt. Default is 10 seconds.
+
+    Returns
+    -------
+    bool
+        True if the user opts in, False otherwise.
+    """
+    print(
+        "PyBaMM can collect usage data and send it to the PyBaMM team to "
+        "help us improve the software.\n"
+        "We do not collect any sensitive information such as models, parameters, "
+        "or simulation results - only information on which parts of the code are "
+        "being used and how frequently.\n"
+        "This is entirely optional and does not impact the functionality of PyBaMM.\n"
+        "For more information, see https://docs.pybamm.org/en/latest/source/user_guide/index.html#telemetry"
+    )
+
+    def get_input():  # pragma: no cover
+        try:
+            user_input = (
+                input("Do you want to enable telemetry? (Y/n): ").strip().lower()
+            )
+            answer.append(user_input)
+        except Exception:
+            # Handle any input errors
+            pass
+
+    time_start = time.time()
+
+    while True:
+        if time.time() - time_start > timeout:
+            print("\nTimeout reached. Defaulting to not enabling telemetry.")
+            return False
+
+        answer = []
+        # Create and start input thread
+        input_thread = threading.Thread(target=get_input)
+        input_thread.daemon = True
+        input_thread.start()
+
+        # Wait for either timeout or input
+        input_thread.join(timeout)
+
+        if answer:
+            if answer[0] in ["yes", "y", ""]:
+                print("\nTelemetry enabled.\n")
+                return True
+            elif answer[0] in ["no", "n"]:
+                print("\nTelemetry disabled.\n")
+                return False
+            else:
+                print("\nInvalid input. Please enter 'yes/y' for yes or 'no/n' for no.")
+        else:
+            print("\nTimeout reached. Defaulting to not enabling telemetry.")
+            return False
+
+
+def generate():
+    if is_running_tests():
+        return
+
+    # Check if the config file already exists
+    if read() is not None:
+        return
+
+    # Ask the user if they want to opt in to telemetry
+    opt_in = ask_user_opt_in()
+    config_file = Path(platformdirs.user_config_dir("pybamm")) / "config.yml"
+    write_uuid_to_file(config_file, opt_in)
+
+    if opt_in:
+        pybamm.telemetry.capture("user-opted-in")
+
+
+def read():
+    config_file = Path(platformdirs.user_config_dir("pybamm")) / "config.yml"
+    return read_uuid_from_file(config_file)
+
+
+def write_uuid_to_file(config_file, opt_in):
+    # Create the directory if it doesn't exist
+    config_file.parent.mkdir(parents=True, exist_ok=True)
+
+    # Write the UUID to the config file in YAML format
+    with open(config_file, "w") as f:
+        f.write("pybamm:\n")
+        f.write(f"  enable_telemetry: {opt_in}\n")
+        if opt_in:
+            unique_id = uuid.uuid4()
+            f.write(f"  uuid: {unique_id}\n")
+
+
+def read_uuid_from_file(config_file):
+    # Check if the config file exists
+    if not config_file.exists():
+        return None
+
+    # Read the UUID from the config file
+    with open(config_file) as f:
+        content = f.read().strip()
+
+    # Extract the UUID using YAML parsing
+    try:
+        import yaml
+
+        config = yaml.safe_load(content)
+        return config["pybamm"]
+    except (yaml.YAMLError, ValueError):
+        return None

src/pybamm/simulation.py

@@ -10,6 +10,7 @@
 import warnings
 from functools import lru_cache
 from datetime import timedelta
+import pybamm.telemetry
 from pybamm.util import import_optional_dependency
 
 from pybamm.expression_tree.operations.serialise import Serialise
@@ -450,6 +451,8 @@ def solve(
             Additional key-word arguments passed to `solver.solve`.
             See :meth:`pybamm.BaseSolver.solve`.
         """
+        pybamm.telemetry.capture("simulation-solved")
+
         # Setup
         if solver is None:
             solver = self._solver

src/pybamm/telemetry.py

@@ -0,0 +1,36 @@
+from posthog import Posthog
+import os
+import pybamm
+import sys
+
+_posthog = Posthog(
+    # this is the public, write only API key, so it's ok to include it here
+    project_api_key="phc_bLZKBW03XjgiRhbWnPsnKPr0iw0z03fA6ZZYjxgW7ej",
+    host="https://us.i.posthog.com",
+)
+
+
+def disable():
+    _posthog.disabled = True
+
+
+_opt_out = os.getenv("PYBAMM_DISABLE_TELEMETRY", "false").lower()
+if _opt_out != "false":  # pragma: no cover
+    disable()
+
+
+def capture(event):  # pragma: no cover
+    # don't capture events in automated testing
+    if pybamm.config.is_running_tests() or _posthog.disabled:
+        return
+
+    properties = {
+        "python_version": sys.version,
+        "pybamm_version": pybamm.__version__,
+    }
+
+    config = pybamm.config.read()
+    if config:
+        if config["enable_telemetry"]:
+            user_id = config["uuid"]
+            _posthog.capture(user_id, event, properties=properties)