### Clone and Setup ROI Reporting Agent

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/agent/README.md

Clone the repository, set up the sparse checkout, create the necessary database, and push the agent. This quick start guide covers the essential steps for initial deployment.

```bash
# 1. Clone repository
git clone --depth 1 --filter=blob:none --sparse https://github.com/treasure-data/treasure-boxes.git
cd treasure-boxes
git sparse-checkout set engage-box/roi_reporting
cd engage-box/roi_reporting/agent

# 2. Create database (if not exists)
tdx api -X POST /v3/database/create/engage_roi_reporting
# Note: "Name has already been taken" error is OK if database already exists

# 3. Create LLM project
tdx llm project create "ROI Reporting Agent"

# 4. Push agent
tdx agent push . -f
```

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

### Start S3 Example Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/s3/README.md

Initiate the S3 example workflow using the `td wf start` command. This command starts a new session for the specified workflow and project.

```sh
$  td wf start s3 s3 --session now
2019-09-14 10:15:23 +0900: Digdag v0.9.39
Started a session attempt:
  session id: 12623047
  attempt id: 55668803
  uuid: 094b3ba3-7b79-4abb-a529-89e5b67ba711
  project: s3
  workflow: s3
  session time: 2019-09-14 01:15:24 +0000
  retry attempt name:
  params: {}
  created at: 2019-09-14 10:15:27 +0900

* Use `td workflow session 12623047` to show session status.
* Use `td workflow task 55668803` and `td workflow log 55668803` to show task status and logs.
```

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

### Python Environment Setup

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/workflows/reporting_agent/README.md

Set up a Python virtual environment and install necessary dependencies using pip. Ensure Python 3.6 or later is installed.

```bash
# Create a virtual environment
python -m venv reporting_agent_env

# Activate the virtual environment
# On macOS/Linux:
source reporting_agent_env/bin/activate
# On Windows (Note: Windows commands are provided for reference but have not been verified):
# reporting_agent_env\Scripts\activate

# Create a requirements.txt file with the following content:
cat > requirements.txt << EOL
pytd>=0.9.0
pandas>=1.0.0
numpy>=1.19.0
EOL

# Install dependencies
pip install -r requirements.txt
```

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

### Push and Start Basic Monitoring Workflow

Source: https://context7.com/treasure-data/treasure-boxes/llms.txt

Use these commands to push the basic monitoring workflow and start initial and incremental ingestions. Ensure secrets are set before starting.

```bash
td wf push basic_monitoring
td wf secrets --project basic_monitoring --set td.apikey=1234/abcdefg...
td wf start basic_monitoring initial_ingest --session now
# After initial load, run incremental daily:
td wf start basic_monitoring incremental_ingest --session now
```

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

### Running td-spark PySpark Example

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/pyspark/README.md

Commands to push the td-spark project, set API keys and endpoint secrets, and start a PySpark session.

```bash
$ td wf push td-spark
$ td wf secrets --project td-spark --set td.apikey --set td.apiserver
# Input TD API KEY (1/XXXXXXX) and ENDPOINT (e.g. https://api.treasuredata.com)
$ td wf start td-spark pyspark --session now
```

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

### Start HTTP Server (Python 3)

Source: https://github.com/treasure-data/treasure-boxes/blob/master/tool-box/workflow-visualizer/README.md

Start a simple HTTP server on port 8000 using Python 3. This is required to serve the visualizer's files locally.

```bash
    $ python -m http.server 8000
```

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

### Start Workflow Session with td wf start

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/power_bi/README.md

Trigger a workflow session manually. Replace `<project_name>` and `<workflow_name>` with your project and workflow names, respectively. 'now' starts the session immediately.

```bash
# Run
    $ td wf start <project_name> <workflow_name> --session now
```

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

### Start HTTP Server (Python 2)

Source: https://github.com/treasure-data/treasure-boxes/blob/master/tool-box/workflow-visualizer/README.md

Start a simple HTTP server on port 8000 using Python 2. This is required to serve the visualizer's files locally.

```bash
    $ python -m SimpleHTTPServer 8000
```

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

### Install Python Packages

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/scorer-edge/README.md

Install necessary Python packages for Treasure Data integration and image processing. Use the --user flag to install packages in the user's home directory.

```sh
pip install --user pytd imutils
```

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

### Start Workflow Session with td wf start

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/backup_and_delete/README.md

Trigger a manual session for the workflow to observe its execution. The '--session now' flag starts the session immediately.

```bash
# Run
$ td wf start backup_and_delete --session now
```

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

### Start Treasure Data Workflow for S3

Source: https://github.com/treasure-data/treasure-boxes/blob/master/tool-box/s3_presigned/README.md

Execute the S3 workflow using `td wf start`. This command initiates the session, which will then generate the pre-signed URL.

```sh
$ td wf start s3_presigned s3 --session now
2019-09-14 10:15:23 +0900: Digdag v0.9.39
Started a session attempt:
  session id: 12623047
  attempt id: 55668803
  uuid: 094b3ba3-7b79-4abb-a529-89e5b67ba711
  project: s3_presigned
  workflow: s3
  session time: 2019-09-14 01:15:24 +0000
  retry attempt name:
  params: {}
  created at: 2019-09-14 10:15:27 +0900

* Use `td workflow session 12623047` to show session status.
* Use `td workflow task 55668803` and `td workflow log 55668803` to show task status and logs.
```

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

### Input JSON Lines Data Example

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/expand_json_s3/README.md

This is an example of the expected input file format, which must be in JSON Lines. Each JSON object should be on a single line.

```json
{"id": 1, "name":"John Smith", "valid_user":true, "email":{"primary": "foo1@example.com", "secondary": "bar1@example.com"}}
{"id": 2, "name":"Joe Bloggs", "valid_user":false, "email":{"primary": "foo2@example.com", "secondary": "bar2@example.com"}}
```

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

### Workflow Log Output Example

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/ml_tips/README.md

Example output from the workflow log showing AUC and Log Loss metrics.

```text
echo>: auc: 0.8044451551007356    logloss: 0.5337748192743182
```

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

### Target List Table Structure Example

Source: https://github.com/treasure-data/treasure-boxes/blob/master/activation_actions/filter_by_target_list/README.md

Provides an example schema for a target list table, which includes anonymous_id for joining.

```sql
campaign_target_list (
  anonymous_id VARCHAR,   -- Anonymous ID (JOIN KEY)
  campaign_id VARCHAR     -- Campaign ID
)
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/python/README.md

Initiates a session for the uploaded workflow. This command starts the execution and provides details for monitoring the session and task status.

```bash
td wf start simple-example simple --session now
2019-05-13 21:04:32 -0700: Digdag v0.9.36
Started a session attempt:
  session id: 250802
  attempt id: 250930
  uuid: 2453cfa8-2b16-4a6f-83c2-9e9c9e7cdb33
  project: simple-example
  workflow: simple
  session time: 2019-05-14 04:04:32 +0000
  retry attempt name: 
  params: {}
  created at: 2019-05-13 21:04:35 -0700

* Use `td workflow session 250802` to show session status.
* Use `td workflow task 250930` and `td workflow log 250930` to show task status and logs.
```

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

### Run Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/auth0_batch/README.md

Start the Auth0 batch workflow for exporting user data.

```bash
$ td wf start auth0_batch export --session now
```

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

### Start Recommendation Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/docs/more.md

Command to start the main recommendation workflow using Treasure Workflow, typically after data preparation.

```bash
$ td wf start recommendation recommend --session now -p apikey={YOUR_API_KEY}
```

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

### Start Workflow Session with td wf start

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/redshift/README.md

Manually trigger a workflow session. Replace `td_redshift` with your workflow's name.

```bash
# Run
$ td wf start sample_project td_redshift --session now
```

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

### List Directory Contents

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/python/README.md

Verifies that you are in the correct directory for the example. This command should show README.md, other_scripts, scripts, and simple.dig.

```bash
$ ls
README.md	other_scripts	scripts	simple.dig
```

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

### Install TD Workflow Wrapper

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/python/README.md

Installs a wrapper script for the TD CLI to simplify workflow management. Ensure you have curl and appropriate permissions to write to /usr/local/bin.

```bash
curl -sL https://gist.github.com/kiyoto/d46ccf71faa924320c0a7993515d0e9c > /usr/local/bin/tdwf
chmod +x /usr/local/bin/tdwf
```

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

### Start Workflow Execution

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/monitoring/workflow_monitoring/README.md

Starts the execution of the 'initial_ingest_session_attempt' workflow for the 'workflow_monitoring' project. The '--session now' argument initiates the workflow immediately.

```bash
$td wf start workflow_monitoring initial_ingest_session_attempt --session now
```

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

### Upload Workflow to TreasureData

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/slack/README.md

Use this command to upload the example workflow to your TreasureData account.

```bash
# Upload
td wf push td_slack_example
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/s3-v2/README.md

Manually trigger a workflow session using the `td wf start` command. Specify the project name and the workflow name.

```bash
# Run
    $ td wf start td_s3_v2 td_s3_v2 --session now
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/collaborative_filtering/README.md

Initiate a new session for the collaborative filtering workflow on Treasure Data.

```bash
td wf start --session now collaborative_filtering dimsum
```

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

### Run Initial Ingest Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/monitoring/basic_monitoring/README.md

Starts the initial ingest workflow for metadata. Use 'now' for the session to run immediately.

```bash
$td wf start basic_monitoring initial_ingest --session now
```

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

### Start Workflow in Server Mode

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_load/gcs/README.md

Manually trigger a workflow session on the server using `td wf start`. This command initiates the `daily_load` session for the `td_load_gcs` workflow.

```bash
# Run
    $ td wf start td_load_gcs daily_load --session now
```

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

### Upload and Run Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_run/README.md

Commands to upload the workflow to Treasure Data and start a session for running saved queries.

```bash
$ td wf push td_run_example
```

```bash
$ td wf start td_run_example run_saved_query --session now
```

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

### Push and Start TD Recommendation Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/README.md

Commands to push the recommendation workflow to Treasure Data and start a recommendation session. Ensure the workflow script and configuration files are present.

```sh
$ ./data.sh
$ td wf push recommendation # push workflow to TD
$ td wf start recommendation recommend --session now
```

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

### Push and Start GDPR Workflow

Source: https://context7.com/treasure-data/treasure-boxes/llms.txt

Commands to push the GDPR workflow and start its execution with specified database and parameters table. This workflow automates GDPR DELETE, UPDATE, and INFORM operations.

```bash
td wf push gdpr_workflow
td wf start gdpr_workflow gdpr_workflow --session now \
  -p database=gdpr_example_db \
  -p params_table=params
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/sftp/README.md

Trigger a workflow session manually using `td wf start`. Specify the project name and workflow name, and optionally a session time like 'now'.

```bash
# Run
    $ td wf start td_sftp td_sftp --session now
```

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

### Upload and Run Treasure Data Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/presto_and_hive/README.md

Commands to upload the workflow to Treasure Data and start a session for the 'nasdaq_analysis' analysis.

```bash
td wf push td_example
```

```bash
td wf start td_example nasdaq_analysis --session now
```

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

### Run Test Setup Script

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/workflows/reporting_agent/README.md

Execute the script to create test tables with sample data. This script requires the database name and can configure the number of revenue samples, days, and realism.

```bash
# From the root of the engage-api repository
python scripts/reporting_agent/setup_test_tables.py --database engage_roi_reporting --create-tables --revenue-samples 500 --days 30 --realistic-revenue
```

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

### Start Workflow Manually

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_load/instagram_user_and_media_insights/README.md

Manually trigger a workflow session. Replace 'daily_load' with your workflow name if different.

```sh
# Run
$ td wf start ig_insights daily_load --session now
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/gcs/README.md

Trigger a workflow session manually using the `td wf start` command. Specify the project name, workflow name, and session time. For example, to start the 'export_gcs' workflow in the 'td_gcs' project for the current time, use '--session now'.

```bash
# Run
$ td wf start td_gcs export_gcs --session now
```

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

### Setup Data Preparation Workflows First

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/README.md

Navigate to the workflows directory to set up the Data Preparation Workflows. This option is suitable for creating tables from scratch.

```bash
cd workflows/reporting_agent/
# Follow instructions in workflows/reporting_agent/README.md
```

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

### Start Treasure Data Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_ddl/README.md

Use this command to start a manual session for the 'ddl_operations' part of the workflow. The '--session now' argument starts it immediately.

```bash
# Run
    $ td wf start td_run_example ddl_operations --session now
```

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

### swimOS Tutorial Application Output

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/swimos/README.md

Example output from the swimOS tutorial application demonstrating data publishing, agent updates, and POST requests to the Treasure Data Postback API.

```text
> Task :run
Running Tutorial plane...
[/unit/master] Hello, world!
[/unit/master] publish: commanded with {foo:60,bar:106,baz:198}
[/unit/master] latest: set to {foo:60,bar:106,baz:198} from
[/unit/master] history: appended {0, {1574293480492,{foo:60,bar:106,baz:198}}}
[/unit/master] histogram: replaced 1574293480000's value to {count:2} from
[/unit/master] Sending POST request to: https://in.treasuredata.com/postback/v3/event/DATABASE_NAME/TABLE_NAME
[/unit/master] Data: {"time":1574293480534,"foo":60,"bar":106,"baz":198}
[/unit/master] Response code: 200
[/unit/master] publish: commanded with {foo:58,bar:134,baz:188}
[/unit/master] latest: set to {foo:58,bar:134,baz:188} from {foo:60,bar:106,baz:198}
[/unit/master] history: appended {1, {1574293481172,{foo:58,bar:134,baz:188}}}
[/unit/master] histogram: replaced 1574293480000's value to {count:21} from {count:2}
[/unit/master] Sending POST request to: https://in.treasuredata.com/postback/v3/event/DATABASE_NAME/TABLE_NAME
[/unit/master] Data: {"time":1574293481172,"foo":58,"bar":134,"baz":188}
[/unit/master] Response code: 200
...

```

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

### Execute ML Tips Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/ml_tips/README.md

Commands to upload the sample dataset, push the workflow, and start a session for the ML tips workflow.

```sh
$ ./data.sh # Upload sample dataset to Treasure Data
$ td wf push ml_tips
$ td wf start ml_tips titanic --session now
```

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

### Create Sample Database

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/elastic-map-reduce/README.md

Create a temporary database in Treasure Data to store the job results. This database may already exist if previous tutorials were run.

```bash
# Sample Database. NOTE: this database may already exist if you've run past tutorials.
$ td db:create workflow_temp
```

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

### Install mbed-cli requirements

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/mbed-os/README.md

Installs the necessary Python packages for mbed-cli.

```sh
pip install -r requirements.txt
```

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

### Setup AI Reporting Agent First

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/README.md

Navigate to the agent directory to set up the AI Reporting Agent. This option is suitable if the database and tables already exist.

```bash
cd agent/
# Follow instructions in agent/README.md
```

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

### Run swimOS Tutorial Application

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/swimos/README.md

Execute the Gradle task to run the swimOS tutorial. This will start a mock data source publishing records to a Web Agent, which then forwards them to Treasure Data.

```sh
./gradlew run
```

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

### Start Treasure Data Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/elasticsearch/README.md

Manually trigger a workflow session using `td wf start`. Specify the project name and the workflow name (e.g., `td_elasticsearch`). The `--session now` argument starts the session immediately.

```bash
td wf start sample_project td_elasticsearch --session now
```

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

### Start Workflow Session Manually

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/result_export/export_result_tableau/README.md

Triggers the workflow session manually. Use 'now' to start the session immediately.

```bash
# Run
$ td wf start export_result_tableau --session now
```

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

### Run Workflow with Existing Authentication

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_load/google_analytics/README.md

Start the `td_load_example` workflow using the `daily_load_with_existing_authentication` task. This method is used when authentication is managed via the Treasure Data console.

```bash
td wf start td_load_example daily_load_with_existing_authentication --session now
```

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

### Prepare Dataset and Push Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/tf-idf/README.md

Execute these commands to prepare the sample dataset, create a stopwords table, push the TF-IDF workflow to Treasure Data, and start a new session.

```sh
$ ./data.sh # prepare sample dataset
$ ./stopwords.sh # prepare stopwords table
$ td wf push tfidf # push workflow to TD
$ td wf start tfidf tfidf --session now -p apikey=${YOUR_TD_API_KEY}
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/nasdaq_analysis/README.md

Initiates a workflow session for 'nasdaq_analysis' starting immediately. This command is used to manually trigger the workflow execution.

```bash
td wf start td_example nasdaq_analysis --session now
```

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

### Prepare Database and Table

Source: https://github.com/treasure-data/treasure-boxes/blob/master/tool-box/get_cdp_segments/README.md

Create the necessary database and table in Treasure Data where the segment data will be imported.

```bash
$ td db:create example_db
$ td table:create master_segment_lists segment_lists
```

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

### Manually Configure Data Source and Target Database

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/docs/more.md

Example of how to manually configure the source table and target database in the `config/params.yml` file.

```yaml
# Data:
source: source_db.table_name # input table
target: target_db # output database
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/twitter_tailored_audience/README.md

Manually trigger a workflow session. This command initiates the execution of your defined workflow. You can specify 'now' to start immediately.

```bash
# Run
    $ td wf start twitter_tailored_audience twitter_tailored_audience --session now
```

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

### Start Treasure Data Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/result_export/export_result_redshift/README.md

Manually trigger a new session for your Treasure Data workflow. Use 'now' to start immediately.

```bash
# Run
$ td wf start export_result_redshift export_result_redshift --session now
```

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

### Navigate to Project Directory

Source: https://github.com/treasure-data/treasure-boxes/blob/master/README.md

Change into the cloned Treasure Boxes directory to begin making changes.

```bash
$ cd treasure-boxes
```

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

### Navigate to Data Preparation Workflows Directory

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/README.md

Change directory to the Data Preparation Workflows component. Detailed setup instructions are available in the workflow's README.

```bash
cd workflows/reporting_agent/
# See workflows/reporting_agent/README.md for detailed setup instructions
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/account_reporting/README.md

Manually trigger a workflow session using the `td wf start` command. Specify the workflow name and session details.

```bash
    # Run
    $ td wf start account_reporting account_reporting --session now
```

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

### Prepare Database and Tables

Source: https://github.com/treasure-data/treasure-boxes/blob/master/data-box/tapad-cross-device-graph/README.md

Create the necessary database and tables in Treasure Data for storing DGA and ID mapping data. Modify settings in tapad_data_ingestion.dig if table names are changed.

```bash
$ td db:create tapad
$ td table:create tapad dga
$ td table:create tapad id_map
$ td table:create tapad idfa
$ td table:create tapad adid
```

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

### Start Recommendation Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/docs/more.md

Initiates the recommendation workflow, including ID mapping if user/item IDs are not long values. Requires your Treasure Data API key and target database.

```bash
$ td wf start recommendation map_id --session now -p apikey=${YOUR_TD_API_KEY}
```

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

### Start Workflow in Local Mode

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_load/gcs/README.md

Manually trigger a workflow session using `td wf start`. This command initiates the `daily_load` session for the `td_load_gcs` workflow.

```bash
    $ td wf start td_load_gcs daily_load --session now
```

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

### Run Marketo Data Ingestion Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td_load/marketo/README.md

Upload the workflow, set the necessary Marketo credentials as secrets, and then start a session for the daily load. Ensure that the target database and tables exist before running.

```bash
# Upload
$ td wf push td_load_example

# Run
$ td wf start td_load_example daily_load --session now
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/postgresql/README.md

Trigger a workflow session manually using the `td wf start` command. This initiates a new run of your Treasure Data workflow.

```bash
# Run
    $ td wf start td_postgresql td_postgresql --session now
```

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

### Start Workflow Session with td

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/ftp/README.md

Trigger a workflow session manually using the `td wf start` command. Specify the workflow name and the session time.

```bash
td wf start td_ftp td_ftp --session now
```

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

### Start Recommendation Prediction Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/recommendation/docs/more.md

Command to start the recommendation prediction workflow using Treasure Workflow. It includes specifying the session and passing an API key.

```bash
$ td wf start recommendation predict --session now -p apikey={YOUR_API_KEY}
```

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

### Initialize Treasure Data SDK

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/datalayer_papi/datalayer_papi.html

Initializes the Treasure Data SDK with essential configuration including the host, write key, and database. It also sets the SDK to start in signed mode.

```javascript
// Initialize TD.
var td = new Treasure({
  host: 'us01.records.in.treasuredata.com', // US RT endpoint
  writeKey: 'xxx/yyyyy', // write-only TD API key
  database: 'zzz' // TD database
  // Send td_ip, td_user_agent, td_path, td_host, td_client_id and td_global_id from the beginning, if set. (Should ask for consent from website visitors first!)
  startInSignedMode: true
});
```

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

### Start Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/kill_wf_attempt/README.md

Manually trigger a workflow session. This command starts the 'kill_wf_attempt' workflow with the 'kill_wf_attempt' session name, set to run immediately ('now').

```bash
# Run
$ td wf start kill_wf_attempt kill_wf_attempt --session now
```

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

### Data Loading Steps

Source: https://github.com/treasure-data/treasure-boxes/blob/master/tool-box/cjo-profile-viewer/docs/PROJECT_SUMMARY.md

Illustrates the two-step process for loading journey configuration and profile data. This involves fetching journey details and then querying profile data based on selected attributes.

```text
Step 1: Load Journey Config → Extract audience ID → Get available attributes
Step 2: Select attributes → Load Profile Data → Enable visualization
```

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

### Start Workflow Session with TD CLI

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/import_iterable_to_treasuredata/README.md

Manually trigger a workflow session using the TD CLI. The '--session now' argument starts an immediate execution.

```bash
# Run
td wf start import_iterable_to_td import_iterable_to_td --session now
```

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

### Start Treasure Data Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/integration-box/scorer-cloud/README.md

Manually start the Treasure Data workflow to process data. This command initiates a session to begin the data ingestion process.

```sh
td wf start scorer scorer --session now
```

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

### Upload Project

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/invalid_v_column/README.md

Use this command to upload the project to Treasure Data.

```bash
td wf push invalid_v_column
```

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

### Top-k Item Recommendation Pipeline Setup

Source: https://context7.com/treasure-data/treasure-boxes/llms.txt

Schema definition for input and output tables for a Matrix Factorization-based collaborative filtering recommendation system. Supports PySpark ALS alternative.

```yaml
# Input table schema (must be named "input")
# userid LONG | itemid LONG | rating FLOAT
# 3077        | 2046        | 4.0

# Output table: recommended items per user
# userid LONG | recommended_items ARRAY<LONG>
# 10          | [4454,2609,1214,1949,33539,7079,2324,224,3989,26939]
```

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

### Run Marketo Daily Load (UpdatedAt Filter - Initialization)

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/marketo_daily_load/README.md

Initiates the Marketo daily load workflow for leads, filtering by their last update timestamp. This is the first step in processing updates. Replace 'YOUR_DATE' with the desired date in 'YYYY-MM-DD' format.

```bash
td wf start marketo_daily_load initialize_updatedat --session 'YOUR_DATE(ex. 2018-01-01)'
```

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

### Run Initial Ingest Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/monitoring/cdp_monitoring/README.md

Starts the initial ingest workflow for CDP metadata. Use 'now' for the session to run immediately.

```bash
td wf start cdp_monitoring initial_ingest --session now
```

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

### List Databases

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/agent/README.md

Lists all available databases. Use this to verify the existence and exact name of the 'engage_roi_reporting' database.

```bash
tdx db list | grep engage_roi_reporting
```

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

### Start Workflow Session Manually with td

Source: https://github.com/treasure-data/treasure-boxes/blob/master/td/mailchimp/README.md

Trigger a workflow session manually using the `td wf start` command, specifying the project and workflow name, and the session time.

```bash
# Run
$ td wf start mailchimp_export mailchimp_export --session now
```

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

### Start Treasure Data Workflow Session

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/result_export/export_result_ftp/README.md

Manually trigger a workflow session using the `td wf start` command. Specify the project, workflow name, and session timing.

```bash
# Run
$ td wf start export_result_ftp export_result_ftp --session now
```

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

### Start Workflow Session with TD CLI

Source: https://github.com/treasure-data/treasure-boxes/blob/master/scenarios/kick_another_account_workflow_kohki/README.md

Manually trigger a workflow session using `td wf start`. This command initiates the execution of the specified workflow in the target account.

```bash
# Run
$ td wf start kick_another_account_workflow_kohki kick_another_account_workflow_kohki --session now
```

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

### Create Database

Source: https://github.com/treasure-data/treasure-boxes/blob/master/engage-box/roi_reporting/agent/README.md

Creates a new database with the specified name. Use this if the 'engage_roi_reporting' database is not found.

```bash
tdx api -X POST /v3/database/create/engage_roi_reporting
```

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

### Push and Start CLTV Prediction Workflow

Source: https://github.com/treasure-data/treasure-boxes/blob/master/machine-learning-box/customer-lifetime-value/README.md

Use these commands to push the CLTV prediction workflow to Treasure Data, set necessary API secrets, and start a prediction session.

```sh
$ td wf push cltv-prediction # push workflow to TD
$ td wf secrets --project cltv-prediction --set td.apikey td.apiserver
$ td wf start cltv-prediction predict --session now 
```