### my_qadapter.yaml Example for SLURM

Source: https://materialsproject.github.io/jobflow/install_fireworks.html

An example my_qadapter.yaml file for the SLURM queuing system. Adjust paths, resource allocations, and pre-execution commands as needed.

```yaml
_fw_name: CommonAdapter
_fw_q_type: SLURM
rocket_launch: rlaunch -w /path/to/fw_config/my_fworker.yaml singleshot
nodes: 2
walltime: 00:30:00
account: <account>
job_name: my_firework
qos: regular
pre_rocket: |
  module load vasp
  export ATOMATE2_VASP_CMD="srun -N 2 --ntasks-per-node=24"

```

--------------------------------

### Install Jobflow from Source

Source: https://materialsproject.github.io/jobflow/install.html

Clone the repository and install the package locally, optionally including extra dependencies or developer mode.

```bash
git clone https://github.com/materialsproject/jobflow.git
cd jobflow
pip install .
```

```bash
pip install .[fireworks]
```

```bash
pip install -e .
```

--------------------------------

### Log Output Example

Source: https://materialsproject.github.io/jobflow/tutorials/2-introduction.html

Example log output from running jobs locally, showing job start and finish times, job names, and unique job UUIDs.

```log
2023-12-08 11:15:18,266 INFO Started executing jobs locally
2023-12-08 11:15:18,270 INFO Starting job - time_website (f7831f60-617e-490d-8854-ef3e25a78504)
2023-12-08 11:15:18,873 INFO Finished job - time_website (f7831f60-617e-490d-8854-ef3e25a78504)
2023-12-08 11:15:18,875 INFO Starting job - time_website (993990cd-872e-4c8f-823b-ac4eea8756c6)
2023-12-08 11:15:19,300 INFO Finished job - time_website (993990cd-872e-4c8f-823b-ac4eea8756c6)
2023-12-08 11:15:19,304 INFO Starting job - time_website (3c2d3abd-be33-4857-9943-6526bfb05804)
2023-12-08 11:15:19,774 INFO Finished job - time_website (3c2d3abd-be33-4857-9943-6526bfb05804)
2023-12-08 11:15:19,777 INFO Starting job - sum_numbers (2be44e87-0918-481f-ae77-7a5931074e1e)
2023-12-08 11:15:19,783 INFO Finished job - sum_numbers (2be44e87-0918-481f-ae77-7a5931074e1e)
2023-12-08 11:15:19,785 INFO Finished executing jobs locally
```

--------------------------------

### Build Documentation

Source: https://materialsproject.github.io/jobflow/install.html

Install documentation dependencies and generate the site using sphinx-build.

```bash
pip install .[docs]
```

```bash
sphinx-build docs/src docs_build
```

--------------------------------

### Install Jobflow via pip

Source: https://materialsproject.github.io/jobflow/index.html

Standard installation command for the Jobflow library.

```bash
pip install jobflow
```

--------------------------------

### Example Jobflow Run Outputs

Source: https://materialsproject.github.io/jobflow/tutorials/2-introduction.html

A concrete example of the 'responses' dictionary returned by run_locally, showing the output for several jobs identified by their UUIDs.

```json
{
    "f7831f60-617e-490d-8854-ef3e25a78504": {
        "1": {
            "output": 0.20485366717912257,
            "detour": null,
            "addition": null,
            "replace": null,
            "stored_data": null,
            "stop_children": false,
            "stop_jobflow": false
        }
    },
    "993990cd-872e-4c8f-823b-ac4eea8756c6": {
        "1": {
            "output": 0.06472958298400044,
            "detour": null,
            "addition": null,
            "replace": null,
            "stored_data": null,
            "stop_children": false,
            "stop_jobflow": false
        }
    },
    "3c2d3abd-be33-4857-9943-6526bfb05804": {
        "1": {
            "output": 0.18873387505300343,
            "detour": null,
            "addition": null,
            "replace": null,
            "stored_data": null,
            "stop_children": false,
            "stop_jobflow": false
        }
    },
    "2be44e87-0918-481f-ae77-7a5931074e1e": {
        "1": {
            "output": 0.45831712521612644,
            "detour": null,
            "addition": null,
            "replace": null,
            "stored_data": null,
            "stop_children": false,
            "stop_jobflow": false
        }
    }
}
```

--------------------------------

### Define FireWorks Configuration Directories

Source: https://materialsproject.github.io/jobflow/tutorials/8-fireworks

Example structure for separate FireWorks configuration directories.

```text
/path/to/fw_config1
├── FW_config1.yaml
├── my_fworker1.yaml
├── my_launchpad1.yaml
└── my_qadapter1.yaml
```

```text
/path/to/fw_config2
├── FW_config2.yaml
├── my_fworker2.yaml
├── my_launchpad2.yaml
└── my_qadapter2.yaml
```

--------------------------------

### Install Jobflow via Pip

Source: https://materialsproject.github.io/jobflow/install.html

Install the core jobflow package or include optional dependencies for specific features.

```bash
pip install jobflow
```

```bash
pip install jobflow[fireworks]
```

```bash
pip install jobflow[vis]  # Install requirements for visualizing jobs and flows
```

--------------------------------

### Configure JobStore with GridFSStore

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/settings.html

Example YAML configuration for a JobStore using MongoStore for documents and GridFSStore for data blobs.

```yaml
docs_store:
  type: MongoStore
  database: jobflow_unittest
  collection_name: outputs
  host: localhost
  port: 27017
additional_stores:
  data:
    type: GridFSStore
    database: jobflow_unittest
    collection_name: outputs_blobs
    host: localhost
    port: 27017
```

--------------------------------

### Run Unit Tests

Source: https://materialsproject.github.io/jobflow/install.html

Install test dependencies and execute the test suite using pytest.

```bash
pip install .[tests]
```

```bash
pytest
```

--------------------------------

### GET /query

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Query the JobStore for multiple documents based on criteria.

```APIDOC
## GET /query

### Description
Query the JobStore for documents matching the provided criteria.

### Method
GET

### Endpoint
/query

### Parameters
#### Query Parameters
- **criteria** (dict) - Optional - PyMongo filter for documents to search in.
- **properties** (dict|list) - Optional - Properties to return in grouped documents.
- **sort** (dict) - Optional - Dictionary of sort order for fields.
- **skip** (int) - Optional - Number of documents to skip.
- **limit** (int) - Optional - Limit on the total number of documents returned.
- **load** (load_type) - Optional - Which items to load from additional stores.

### Response
#### Success Response (200)
- **Iterator[dict]** - The documents matching the query.
```

--------------------------------

### GET /as_dict

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/reference.html

Serializes the current OutputReference instance into a dictionary format.

```APIDOC
## GET /as_dict

### Description
Serialize the reference as a dict.

### Method
GET

### Response
#### Success Response (200)
- **schema** (object) - The output schema associated with the reference.
```

--------------------------------

### Example: Using Flow Output in Another Flow

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/flow.html

Shows how to use the output of a previously defined Flow as an input for a new job in an outer Flow. This highlights the composability of Flows.

```python
>>> add_third = add(flow.output, 5)
>>> outer_flow = Flow(jobs=[flow, add_third])
```

--------------------------------

### Example: Connected Add Jobs in a Flow

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/flow.html

Illustrates creating a Flow with two connected 'add' jobs. The output of the first job is used as an input for the second, demonstrating data dependency within a Flow.

```python
>>> from jobflow import job, Flow
>>> @job
... def add(a, b):
...     return a + b
>>> add_first = add(1, 2)
>>> add_second = add(add_first.output, 2)
>>> flow = Flow(jobs=[add_first, add_second])
```

--------------------------------

### Run Job with JobStore

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/job.html

Executes the job using a provided JobStore. Resolves OutputReferences if configured and logs job start information. Imports necessary modules like `types`, `datetime`, `jobflow`.

```python
import types
from datetime import datetime

from jobflow import CURRENT_JOB
from jobflow.core.flow import get_flow
from jobflow.core.schemas import JobStoreDocument

index_str = f", {self.index}" if self.index != 1 else ""
logger.info(f"Starting job - {self.name} ({self.uuid}{index_str})")
CURRENT_JOB.job = self

if self.config.expose_store:
    CURRENT_JOB.store = store

if self.config.resolve_references:
```

--------------------------------

### Set FW_CONFIG_FILE Environment Variable

Source: https://materialsproject.github.io/jobflow/install_fireworks.html

Example of how to set the FW_CONFIG_FILE environment variable in your ~/.bashrc file to point to your FW_config.yaml.

```bash
export FW_CONFIG_FILE="/path/to/config/fw_config/FW_config.yaml"

```

--------------------------------

### Local Execution Log Output

Source: https://materialsproject.github.io/jobflow/tutorials/5-dynamic-flows.html

Example log output generated during the local execution of fibonacci jobs.

```text
2023-11-23 23:55:13,324 INFO Started executing jobs locally
2023-11-23 23:55:13,325 INFO Starting job - fibonacci (6f7a1e92-577f-4289-85e5-e07a82e542db)
2023-11-23 23:55:13,329 INFO Finished job - fibonacci (6f7a1e92-577f-4289-85e5-e07a82e542db)
2023-11-23 23:55:13,329 INFO Starting job - fibonacci (862ce31a-a963-419a-9862-f419c47c3c9e)
2023-11-23 23:55:13,332 INFO Finished job - fibonacci (862ce31a-a963-419a-9862-f419c47c3c9e)
2023-11-23 23:55:13,333 INFO Starting job - fibonacci (81c28a0a-f1ad-487c-b484-46729f6ca9a6)
2023-11-23 23:55:13,335 INFO Finished job - fibonacci (81c28a0a-f1ad-487c-b484-46729f6ca9a6)
2023-11-23 23:55:13,336 INFO Starting job - fibonacci (084e2022-5a00-4ca8-9edf-46977ca08388)
2023-11-23 23:55:13,338 INFO Finished job - fibonacci (084e2022-5a00-4ca8-9edf-46977ca08388)
2023-11-23 23:55:13,339 INFO Starting job - fibonacci (382f371c-9e92-4185-8dbb-829801ec1afd)
2023-11-23 23:55:13,342 INFO Finished job - fibonacci (382f371c-9e92-4185-8dbb-829801ec1afd)
2023-11-23 23:55:13,342 INFO Starting job - fibonacci (f4521225-993a-4b42-9818-8aaff0da9181)
2023-11-23 23:55:13,345 INFO Finished job - fibonacci (f4521225-993a-4b42-9818-8aaff0da9181)
2023-11-23 23:55:13,346 INFO Starting job - fibonacci (3d475e3e-3774-4015-b056-a23f5cf1da6a)
2023-11-23 23:55:13,349 INFO Finished job - fibonacci (3d475e3e-3774-4015-b056-a23f5cf1da6a)
2023-11-23 23:55:13,351 INFO Starting job - fibonacci (ddc61891-1ce1-4d96-b66e-cb60d52e642a)
2023-11-23 23:55:13,353 INFO Finished job - fibonacci (ddc61891-1ce1-4d96-b66e-cb60d52e642a)
2023-11-23 23:55:13,354 INFO Starting job - fibonacci (499ab638-d8cc-4272-b2b4-f64d246f3676)
2023-11-23 23:55:13,357 INFO Finished job - fibonacci (499ab638-d8cc-4272-b2b4-f64d246f3676)
2023-11-23 23:55:13,357 INFO Starting job - fibonacci (71b7d00f-0109-4a23-8a54-fb29ef143cd2)
2023-11-23 23:55:13,360 INFO Finished job - fibonacci (71b7d00f-0109-4a23-8a54-fb29ef143cd2)
2023-11-23 23:55:13,361 INFO Starting job - fibonacci (d60f3f5d-821a-434a-a8c1-50468b381d5f)
2023-11-23 23:55:13,363 INFO Finished job - fibonacci (d60f3f5d-821a-434a-a8c1-50468b381d5f)
2023-11-23 23:55:13,364 INFO Starting job - fibonacci (7828008e-2f89-4226-aaac-66f25d4ccf0a)
2023-11-23 23:55:13,367 INFO Finished job - fibonacci (7828008e-2f89-4226-aaac-66f25d4ccf0a)
2023-11-23 23:55:13,367 INFO Starting job - fibonacci (062b0cad-2b62-4f42-bd39-ac1b3c90d141)
2023-11-23 23:55:13,370 INFO Finished job - fibonacci (062b0cad-2b62-4f42-bd39-ac1b3c90d141)
2023-11-23 23:55:13,371 INFO Starting job - fibonacci (6c1bf8e4-1808-42b4-a99e-49dd031894bd)
2023-11-23 23:55:13,373 INFO Finished job - fibonacci (6c1bf8e4-1808-42b4-a99e-49dd031894bd)
2023-11-23 23:55:13,374 INFO Starting job - fibonacci (d82e352d-1ca3-40d2-85d1-3f08989d3551)
2023-11-23 23:55:13,376 INFO Finished job - fibonacci (d82e352d-1ca3-40d2-85d1-3f08989d3551)
2023-11-23 23:55:13,377 INFO Finished executing jobs locally
```

--------------------------------

### Configure JobStore with S3Store

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/settings.html

Example YAML configuration for a JobStore using S3Store for data blobs, requiring the key field to be set to blob_uuid.

```yaml
docs_store:
  type: MongoStore
  database: jobflow_unittest
  collection_name: outputs
  host: localhost
  port: 27017
additional_stores:
  data:
    type: S3Store
    bucket: output_blobs
    key: blob_uuid
    index:
      type: MongoStore
      database: jobflow_unittest
      collection_name: output_blobs_index
      host: localhost
      port: 27017
      key: blob_uuid
```

--------------------------------

### Example Output for Code 1 Execution

Source: https://materialsproject.github.io/jobflow/tutorials/7-generalized-makers.html

This output shows the execution log when running a flow generated by Code1Maker, demonstrating the independent and code-specific steps.

```log
2023-06-08 09:57:53,829 INFO Started executing jobs locally
2023-06-08 09:57:53,896 INFO Starting job - job1 (b6557183-838d-4c04-b1ab-356b8923016d)
DO CODE INDEPENDENT STUFF
DO STUFF specific to CODE 1
2023-06-08 09:57:53,896 INFO Finished job - job1 (b6557183-838d-4c04-b1ab-356b8923016d)
2023-06-08 09:57:53,896 INFO Finished executing jobs locally
```

--------------------------------

### Define Add and Conditional Add Jobs (Output > 10)

Source: https://materialsproject.github.io/jobflow/tutorials/5-dynamic-flows.html

Similar to the previous example, this defines `add` and `add_with_logic` jobs. However, it sets up a scenario where the initial addition results in a value greater than 10, demonstrating that `Response(addition=...)` will not trigger a new job in this case. Requires `@job` and `Response`.

```python
@job
def add(a, b):
    return a + b


@job
def add_with_logic(a, b):
    if a < 10:
        return Response(addition=add(a, b))
    return None


job1 = add(1, 20)
job2 = add_with_logic(job1.output, 20)
flow = Flow([job1, job2])

responses = run_locally(flow)

```

--------------------------------

### Initialize DictMods Class

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/utils/dict_mods.html

Initializes the DictMods class by populating a dictionary of supported actions based on the class's static methods. This setup is performed once upon instantiation.

```python
class DictMods:
    """
    Class to define mongo-like modifications on a dict.

    Supported keywords include the following Mongo-based keywords, with the usual
    meanings (refer to Mongo documentation for information):

    - ``_inc``
    - ``_set``
    - ``_unset``
    - ``_push``
    - ``_push_all``
    - ``_add_to_set`` (but ``_each`` is not supported)
    - ``_pop``
    - ``_pull``
    - ``_pull_all``
    - ``_rename``

    .. Note::

        Note that ``_set`` does not support modification of nested dicts using the mongo
        ``{"a.b":1}`` notation. This is because mongo does not allow keys with "." to be
        inserted. Instead, nested dict modification is supported using a special "->"
        keyword, e.g. ``{"a->b": 1}``
    """

    def __init__(self):
        self.supported_actions = {}
        for i in dir(self):
            if (not re.match(r"__\w+__", i)) and callable(getattr(self, i)):
                self.supported_actions["_{}".format(i)] = getattr(self, i)
```

--------------------------------

### Define a simple job

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Defines a basic job function that performs addition. This serves as a starting point for demonstrating configuration updates.

```python
>>> from jobflow import job, JobConfig
>>> @job
... def add(a, b):
...     return a + b
>>> add_job = add(1, 2)
```

--------------------------------

### Example Output for Code 2 Execution

Source: https://materialsproject.github.io/jobflow/tutorials/7-generalized-makers.html

This output logs the execution of a flow created by Code2Maker, illustrating the sequence of independent and code-specific operations.

```log
2023-06-08 09:57:53,900 INFO Started executing jobs locally
2023-06-08 09:57:53,900 INFO Starting job - job1 (7d744b4f-e0ab-46f0-adb2-72371e0c5070)
DO CODE INDEPENDENT STUFF
DO STUFF specific to CODE 2
2023-06-08 09:57:53,901 INFO Finished job - job1 (7d744b4f-e0ab-46f0-adb2-72371e0c5070)
2023-06-08 09:57:53,901 INFO Finished executing jobs locally
```

--------------------------------

### JobStore Configuration from File

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Demonstrates how to load a JobStore configuration from a file, which can specify different storage backends for documents and additional data.

```APIDOC
## JobStore Configuration from File

### Description
This method loads a JobStore configuration from a file. The file can specify a primary document store (e.g., MongoStore) and additional stores for data (e.g., GridFSStore).

### Method
`classmethod`

### Endpoint
N/A (Class Method)

### Parameters
#### Path Parameters
- **db_file** (str) - Required - Path to the file containing the credentials.

#### Query Parameters
None

#### Request Body
None

### Request Example
None

### Response
#### Success Response (JobStore)
- **JobStore** - An instance of the JobStore class configured with the specified storage.

#### Response Example
None

### Notes
- The `db_file` can contain a serialized JobStore object or a dictionary specification for the stores.
- Additional keyword arguments can be passed to the JobStore constructor.
```

--------------------------------

### Initialize JobStore from File

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Creates a JobStore instance from a JSON or YAML configuration file.

```python
@classmethod
def from_file(cls, db_file: str | Path, **kwargs) -> Self:
    """
    Create a JobStore from a database file.

    Two options are supported for the database file. The file should be in json or
    yaml format.

    The simplest format is a monty dumped version of the store, generated using:

    >>> from monty.serialization import dumpfn
    >>> dumpfn(job_store, "job_store.yaml")

    Alternatively, the file can contain the keys docs_store, additional_stores and
    any other keyword arguments supported by the :obj:`JobStore` constructor.
    """
```

```yaml
docs_store:
```

--------------------------------

### JobStore Configuration from Dictionary Specification

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Illustrates creating a JobStore instance directly from a dictionary that defines the storage configurations.

```APIDOC
## JobStore Configuration from Dictionary Specification

### Description
This class method creates a JobStore from a dictionary specification. The dictionary must include a `docs_store` key and can optionally include `additional_stores` and other keyword arguments for the JobStore constructor. Each store definition requires a `type` key matching a Maggma Store subclass.

### Method
`classmethod`

### Endpoint
N/A (Class Method)

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **spec** (dict) - Required - The dictionary specification for the JobStore configuration.
  - **docs_store** (dict) - Required - Configuration for the primary document store.
    - **type** (str) - Required - The type of the store (e.g., `MongoStore`).
    - **database** (str) - Required - The database name.
    - **collection_name** (str) - Required - The collection name.
    - **host** (str) - Required - The host address.
    - **port** (int) - Required - The port number.
  - **additional_stores** (dict) - Optional - A dictionary of additional stores.
    - **[store_name]** (dict) - Required - Configuration for an additional store.
      - **type** (str) - Required - The type of the store (e.g., `GridFSStore`).
      - **database** (str) - Required - The database name.
      - **collection_name** (str) - Required - The collection name.
      - **host** (str) - Required - The host address.
      - **port** (int) - Required - The port number.
- **kwargs** - Optional - Additional keyword arguments passed to the JobStore constructor.

### Request Example
```yaml
docs_store:
  type: MongoStore
  database: jobflow_unittest
  collection_name: outputs
  host: localhost
  port: 27017
additional_stores:
  data:
    type: GridFSStore
    database: jobflow_unittest
    collection_name: outputs_blobs
    host: localhost
    port: 27017
```

### Response
#### Success Response (JobStore)
- **JobStore** - An instance of the JobStore class configured with the specified storage.

#### Response Example
None

### Errors
- **ValueError**: If `docs_store` is not present in the `spec` dictionary.
```

--------------------------------

### Initialize JobStore

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

The JobStore constructor requires a primary docs_store and accepts optional additional stores, save configurations, and load settings.

```python
def __init__(
    self,
    docs_store: Store,
    additional_stores: dict[str, Store] = None,
    save: save_type = None,
    load: load_type = False,
):
```

--------------------------------

### GET /query_one

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Query the JobStore for a single document.

```APIDOC
## GET /query_one

### Description
Query the Store for a single document matching the criteria.

### Method
GET

### Endpoint
/query_one

### Parameters
#### Query Parameters
- **criteria** (dict) - Optional - PyMongo filter for documents to search.
- **properties** (dict|list) - Optional - Properties to return in the document.
- **sort** (dict) - Optional - Dictionary of sort order for fields.
- **load** (load_type) - Optional - Which items to load from additional stores.

### Response
#### Success Response (200)
- **dict|None** - The document if found, otherwise None.
```

--------------------------------

### GET /flow/get_flow

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Validates dependencies and returns a Flow object.

```APIDOC
## GET /flow/get_flow

### Description
Check dependencies and return a flow object.

### Parameters
#### Query Parameters
- **flow** (Flow | Job | list[jobflow.Job]) - Required - A job, list of jobs, or flow.
- **allow_external_references** (bool) - Optional - If False all the references to other outputs should be from other Jobs of the Flow.

### Response
#### Success Response (200)
- **Flow** (Flow) - A Flow object where connections have been checked.
```

--------------------------------

### Create a Job with a Standard Library Function

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Demonstrates initializing a Job using a function from the Python standard library.

```python
>>> import os
>>> Job(function=os.path.join, function_args=("folder", "filename.txt"))
```

--------------------------------

### Initialize JobStore with MemoryStore

Source: https://materialsproject.github.io/jobflow/tutorials/2-introduction.html

Initialize the JobStore, passing the configured MemoryStore instance. This JobStore will manage the storage and retrieval of job outputs.

```python
from jobflow import JobStore

store = JobStore(docs_store)
```

--------------------------------

### Draw a networkx graph

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/utils/graph.html

Visualizes a graph using matplotlib. Requires matplotlib to be installed.

```python
@requires(mpl, "matplotlib must be installed to plot flow graphs.")
def draw_graph(
    graph: nx.DiGraph,
    layout_function: typing.Callable = None,
    figsize: tuple[float, float] = (12, 8),
):
    """
    Draw a networkx graph.

    Parameters
    ----------
    graph
        A graph object.
    layout_function
        A networkx layout function to use as the graph layout. For example,
        :obj:`.planar_layout`.
    figsize
        The figure size as a tuple of ``(width, height)``.

    Returns
    -------
    matplotlib.pyplot
        The matplotlib pyplot object.
    """
    import matplotlib.pyplot as plt

    if layout_function is None:
        try:
            pos = nx.nx_pydot.graphviz_layout(graph, prog="dot")
        except (ImportError, FileNotFoundError):
            # graphviz not installed
            pos = nx.planar_layout(graph)
    else:
        pos = layout_function(graph)

    plt.figure(figsize=figsize)

    nodes = graph.nodes()
    labels = nx.get_node_attributes(graph, "label")

    nx.draw_networkx_edges(graph, pos)
    nx.draw_networkx_nodes(
        graph, pos, nodelist=nodes, node_color="#B65555", linewidths=1, edgecolors="k"
    )
    nx.draw_networkx_labels(graph, pos, labels=labels)

    edge_labels = nx.get_edge_attributes(graph, "properties")
    nx.draw_networkx_edge_labels(graph, pos, edge_labels=edge_labels, rotate=False)

    plt.gca().axis("off")

    return plt
```

--------------------------------

### Connect to Data Sources

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Establishes connections to the primary docs_store and all configured additional stores.

```python
def connect(self, force_reset: bool = False):
    """
    Connect to the source data.

    Parameters
    ----------
    force_reset
        Whether to reset the connection or not.
    """
    self.docs_store.connect(force_reset=force_reset)
    for additional_store in self.additional_stores.values():
        additional_store.connect(force_reset=force_reset)
```

--------------------------------

### GET /groupby

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Groups documents by specified keys with optional filtering, sorting, and pagination.

```APIDOC
## GET /groupby

### Description
Group documents by keys and return them as an iterator of (key, documents) tuples.

### Method
GET

### Parameters
#### Query Parameters
- **keys** (list[str] | str) - Required - Fields to group documents by.
- **criteria** (dict) - Optional - PyMongo filter for documents to search.
- **properties** (dict | list) - Optional - Properties to return in grouped documents.
- **sort** (dict) - Optional - Dictionary of sort order for fields.
- **skip** (int) - Optional - Number of documents to skip.
- **limit** (int) - Optional - Limit on the total number of documents returned.
- **load** (any) - Optional - Items to load from additional stores.

### Response
#### Success Response (200)
- **data** (list) - An iterator of tuples containing the grouping key dictionary and the list of documents.
```

--------------------------------

### Create a Job with a Custom Function

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Demonstrates initializing a Job using a custom, importable function.

```python
>>> def add(a, b):
...     return a + b
>>> add_job = Job(function=add, function_args=(1, 2))
```

--------------------------------

### Draw Flow Graph

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/flow.html

Draws the flow graph using matplotlib. Requires matplotlib to be installed.

```APIDOC
## POST /draw_graph

### Description
Draws the flow graph using matplotlib. Requires matplotlib to be installed.

### Method
POST

### Endpoint
/draw_graph

### Parameters
#### Query Parameters
- **kwargs** (dict) - Keyword arguments passed to `jobflow.utils.graph.draw_graph`.

### Response
#### Success Response (200)
- **pyplot** (pyplot) - The matplotlib pyplot state object.

#### Response Example
```json
{
  "pyplot": "<matplotlib.pyplot state object>"
}
```
```

--------------------------------

### Create a Job with a Builtin Function

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Demonstrates initializing a Job using a Python builtin function.

```python
>>> print_task = Job(function=print, function_args=("I am a job", ))
```

--------------------------------

### Visualize Flow Graph

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/flow.html

Renders the flow graph using matplotlib. Requires matplotlib to be installed.

```python
def draw_graph(self, **kwargs):
    """
    Draw the flow graph using matplotlib.

    Requires matplotlib to be installed.

    Parameters
    ----------
    kwargs
        keyword arguments that are passed to :obj:`jobflow.utils.graph.draw_graph`.

    Returns
    -------
    pyplot
        The matplotlib pyplot state object.
    """
    from jobflow.utils.graph import draw_graph

    return draw_graph(self.graph, **kwargs)
```

--------------------------------

### POST /from_file

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Creates a JobStore instance from a JSON or YAML configuration file.

```APIDOC
## POST /from_file

### Description
Create a JobStore from a database file in JSON or YAML format.

### Method
POST

### Parameters
#### Request Body
- **db_file** (str) - Required - Path to the database configuration file.
- **kwargs** (dict) - Optional - Additional keyword arguments for the JobStore constructor.
```

--------------------------------

### Create JobStore from File

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Loads a JobStore configuration from a file. If the file contains a serialized JobStore object, it's loaded directly. Otherwise, it's interpreted as a dictionary specification.

```python
from monty.serialization import loadfn

store_info = loadfn(db_file)

if isinstance(store_info, JobStore):
    return store_info

return cls.from_dict_spec(store_info, **kwargs)
```

--------------------------------

### GET /job/serialize

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/job.html

Serializes the current job state into a dictionary format suitable for storage or transmission.

```APIDOC
## GET /job/serialize

### Description
Serializes the job object into a dictionary, ensuring that non-serializable elements like functions or classes are handled via monty/jsanitize.

### Method
GET

### Endpoint
/job/serialize

### Response
#### Success Response (200)
- **data** (dict) - The serialized representation of the job.
```

--------------------------------

### GET /get_output

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Retrieves the output of a specific job UUID, automatically resolving any internal references.

```APIDOC
## GET /get_output

### Description
Get the output of a job UUID. This function automatically resolves any output references in the job outputs.

### Method
GET

### Parameters
#### Query Parameters
- **uuid** (str) - Required - A job UUID.
- **which** (str | int) - Optional - Which index to use: "last" (default), "first", "all", or an integer index.
- **load** (bool) - Optional - Whether to load items from additional stores.
- **cache** (dict) - Optional - A dictionary cache to use for resolving reference values.
- **on_missing** (str) - Optional - Strategy for handling unresolved references.

### Response
#### Success Response (200)
- **output** (Any) - The output(s) for the job UUID.
```

--------------------------------

### Get Job Input UUIDs

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/job.html

Extracts the UUIDs from all OutputReference objects found in the job's inputs.

```python
return tuple(ref.uuid for ref in self.input_references)
```

--------------------------------

### Define and execute a simple Jobflow workflow

Source: https://materialsproject.github.io/jobflow/index.html

Demonstrates how to use the @job decorator to create jobs and combine them into a Flow with automatic dependency resolution.

```python
from jobflow import job, Flow

@job
def add(a, b):
    return a + b

add_first = add(1, 5)
add_second = add(add_first.output, 5)

flow = Flow([add_first, add_second])
flow.draw_graph().show()
```

--------------------------------

### my_launchpad.yaml Template for MongoDB

Source: https://materialsproject.github.io/jobflow/install_fireworks.html

Template for my_launchpad.yaml when using a standard MongoDB connection. Replace placeholders with your database credentials.

```yaml
host: <host name>
port: 27017
name: <database name>
username: <username>
password: <password>
logdir: null
Istrm_lvl: DEBUG
user_indices: []
wf_user_indices: []

```

--------------------------------

### GET /get_output

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Retrieves the output of a specific job UUID, with options to resolve references and handle multiple runs.

```APIDOC
## GET /get_output

### Description
Get the output of a job UUID. This function will automatically try to resolve any output references in the job outputs.

### Method
GET

### Parameters
#### Query Parameters
- **uuid** (str) - Required - A job UUID.
- **which** (str|int) - Optional - Which index to use if there are multiple job runs (default: "last").
- **load** (load_type) - Optional - Which items to load from additional stores.
- **cache** (dict) - Optional - A dictionary cache to use for resolving of reference values.
- **on_missing** (OnMissing) - Optional - What to do if the output contains a reference and the reference cannot be resolved.

### Response
#### Success Response (200)
- **output** (Any) - The output(s) for the job UUID.
```

--------------------------------

### Customizing Nested Maker

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/maker.html

Shows how to customize a nested Maker (e.g., 'add_maker') by providing a custom instance when initializing the parent Maker.

```python
from dataclasses import dataclass
from jobflow import Maker, job

@dataclass
class AddMaker(Maker):
    name: str = "add"
    c: float = 1.5

    @job
    def make(self, a, b):
        return a + b + self.c

@dataclass
class SquareThenAddMaker(Maker):
    name: str = "square then add"
    add_maker: Maker = AddMaker()

    @job
    def make(self, a, b):
        a_squared = a ** 2
        b_squared = b ** 2
        add_job = self.add_maker.make(a_squared, b_squared)
        return Response(detour=add_job)

custom_add_maker = AddMaker(c=1000)
maker = SquareThenAddMaker(add_maker=custom_add_maker)
```

--------------------------------

### Create and Visualize a Jobflow

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/utils/graph.html

Define jobs, chain them, and generate mermaid syntax for visualization. Use mermaid.live to render the graph.

```python
from jobflow import job, Flow

@job
def add(a, b):
    return a + b

add_first = add(1, 2)
add_second = add(add_first.output, 2)
my_flow = Flow(jobs=[add_first, add_second])
graph_source = to_mermaid(my_flow)
```

--------------------------------

### Get Job Input References

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/job.html

Retrieves all OutputReference objects from a job's inputs. Requires the `find_and_get_references` function.

```python
from jobflow.core.reference import find_and_get_references

references: set[jobflow.OutputReference] = set()
for arg in tuple(self.function_args) + tuple(self.function_kwargs.values()):
    references.update(find_and_get_references(arg))

return tuple(references)
```

--------------------------------

### Annotated Log Message Example

Source: https://materialsproject.github.io/jobflow/tutorials/2-introduction.html

An annotated log message illustrating the components: timestamp, job name, and job UUID.

```log
2021-06-04 17:36:50,531 INFO Starting job - time_website (d5460743-bb88-4922-8306-0400eb419a91)
                ↑                                 ↑                         ↑
              time                             job name                  job uuid
```

--------------------------------

### Serialize JobStore to file

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Use monty to dump a JobStore instance into a YAML file.

```python
>>> from monty.serialization import dumpfn
>>> dumpfn(job_store, "job_store.yaml")
```

--------------------------------

### Define and execute a simple workflow

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Create a workflow by connecting job outputs and executing it locally.

```python
>>> from jobflow import job, Flow
>>> @job
... def add(a, b):
...     return a + b
>>> add_first = add(1, 2)
>>> add_second = add(add_first.output, 2)
>>> flow = Flow(jobs=[add_first, add_second])
```

```python
>>> flow = Flow(jobs=[add_first, add_second], output=add_second.output)
```

```python
>>> add_third = add(flow.output, 5)
>>> outer_flow = Flow(jobs=[flow, add_third])
```

```python
>>> from jobflow.managers.local import run_locally
>>> response = run_locally(flow)
```

--------------------------------

### Maker with Instance Variables

Source: https://materialsproject.github.io/jobflow/jobflow.core.html

Makers can have class instance variables that are used within the make function. This example adds a 'c' value to the sum.

```python
@dataclass
class AddMaker(Maker):
    name: str = "add"
    c: float = 1.5

    @job
    def make(self, a, b):
        return a + b + self.c
```

```python
maker = AddMaker(c=3.5)
add_job = maker.make(1, 2)
```

--------------------------------

### JobStore Initialization

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/store.html

Initializes a JobStore instance with a primary document store and optional additional stores for large data. It enforces specific keying strategies for efficient data management.

```APIDOC
## JobStore

### Description

Store intended to allow pushing and pulling documents into multiple stores.

### Parameters

#### Path Parameters
None

#### Query Parameters
None

#### Request Body
None

### `__init__` Method

```python
JobStore(
    docs_store: Store,
    additional_stores: dict[str, Store] = None,
    save: save_type = None,
    load: load_type = False,
)
```

#### Parameters

- **docs_store** (Store) - Required - Store for basic documents.
- **additional_stores** (dict[str, Store]) - Optional - Additional stores to use for storing large data/specific objects. Given as a mapping of ``{store name: store}`` where store is a maggma :obj:`.Store` object.
- **save** (save_type) - Optional - Which items to save in additional stores when uploading documents. Given as a mapping of ``{store name: store_type}`` where ``store_type`` can be a dictionary key (string or enum), an :obj:`.MSONable` class, or a list of keys/classes. If the list of keys/classes itself contains a list/tuple, this will be treated as the path to the save key. Can be used if a key is duplicate in the output and only a single occurrence shall be put in the additional store.
- **load** (load_type) - Optional - Which items to load from additional stores when querying documents. Given as a mapping of ``{store name: store_type}`` where ``store_type`` can be `True``, in which case all saved items are loaded, a dictionary key (string or enum), an :obj:`.MSONable` class, or a list of keys/classes. Alternatively, ``load=True`` will automatically load all items from every additional store.

### Request Example
```python
from maggma.stores import MemoryStore
from jobflow.core.store import JobStore

docs_store = MemoryStore("my_db")
additional_store = MemoryStore("my_large_data_store")

job_store = JobStore(
    docs_store=docs_store,
    additional_stores={"large_data": additional_store},
    save={"data": "my_custom_class"},
    load=True
)
```

### Response
None (initializes the object)

```

--------------------------------

### Get Maker Object from Job

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/job.html

Retrieves the host Maker object if the job is configured to run a maker function. Imports `Maker` from `jobflow`.

```python
from jobflow import Maker

bound = getattr(self.function, "__self__", None)
if isinstance(bound, Maker):
    return bound
return None
```

--------------------------------

### my_launchpad.yaml Template for MongoDB URI

Source: https://materialsproject.github.io/jobflow/install_fireworks.html

Template for my_launchpad.yaml when connecting via a MongoDB URI, such as with MongoDB Atlas. Set uri_store to true.

```yaml
host: <URI>
port: 27017
name: <database name>
uri_store: true
logdir: null
Istrm_lvl: DEBUG
user_indices: []
wf_user_indices: []

```

--------------------------------

### Initialize MemoryStore for Job Outputs

Source: https://materialsproject.github.io/jobflow/tutorials/2-introduction.html

Create an instance of MemoryStore to hold job outputs in memory. This is suitable for testing or development; use permanent stores for production.

```python
from maggma.stores import MemoryStore

docs_store = MemoryStore()
```

--------------------------------

### Initialize FireWorks Launchpad

Source: https://materialsproject.github.io/jobflow/install_fireworks.html

Command to reset and initialize the FireWorks launchpad, ensuring a connection to the database. Use with caution as it clears existing data.

```bash
lpad reset

```

--------------------------------

### Get Flow

Source: https://materialsproject.github.io/jobflow/_modules/jobflow/core/flow.html

Validates dependencies and returns a Flow object. It ensures all job references within the flow are resolvable, either internally or externally if allowed.

```APIDOC
## GET /api/flow

### Description
Checks dependencies and returns a flow object. Ensures that all references to other outputs are either from other Jobs within the Flow or, if allowed, from external sources.

### Method
GET

### Endpoint
/api/flow

### Parameters
#### Query Parameters
- **flow** (Flow | Job | list[jobflow.Job]) - Required - The job, list of jobs, or flow to be checked.
- **allow_external_references** (bool) - Optional - Defaults to False. If False, all references to other outputs must be from other Jobs of the Flow.

### Response
#### Success Response (200)
- **Flow** (Flow) - A :obj:`Flow` object where connections have been checked.

#### Response Example
```json
{
  "flow_object": "..."
}
```
```

--------------------------------

### Demonstrate Response(detour) in JobFlow

Source: https://materialsproject.github.io/jobflow/tutorials/5-dynamic-flows.html

Use Response(detour) to switch to a parallel job or flow. This example shows how a conditional detour to an 'add' job is triggered.

```python
@job
def add(a, b):
    return a + b


@job
def add_with_logic(a, b):
    if a < 10:
        return Response(detour=add(a, b))
    return None


job1 = add(1, 2)
job2 = add_with_logic(job1.output, 2)
flow = Flow([job1, job2])

responses = run_locally(flow)

```