[MRG] fix prevewing public dataset errors

[HuaizhengZhang]

User description

Closes #188

---

PR Type

enhancement, bug fix

---

Description

• Clarified the conditions under which the preview_csv_data function should be used for public datasets in the advisor documentation.
• Changed the default mode in the CLI from 'general' to 'baseline' to better align with the intended functionality.
• Updated the error message in the CLI to reflect the new mode options, ensuring users are informed of valid modes.

---

Changes walkthrough 📝

	Relevant files
Documentation	advisor.py Clarify dataset preview conditions in advisor documentation mle/agents/advisor.py Clarified the usage of preview_csv_data for public datasets. Updated documentation to specify conditions for previewing datasets. +1/-1
Enhancement	cli.py Update CLI default mode and error message                                mle/cli.py Changed default mode from 'general' to 'baseline'. Updated error message to reflect new mode options. +3/-3

---

💡 PR-Agent usage:
Comment /help on the PR to get a list of all available PR-Agent tools and their descriptions

[github-actions]

PR Reviewer Guide 🔍

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Key issues to review Ambiguity in Documentation The added documentation in line 58 is ambiguous. It's unclear whether the function preview_csv_data should or should not be used for public datasets. Consider rephrasing for clarity. Error Handling The error message in line 59 only specifies 'baseline' and 'report' as valid modes. Ensure this aligns with all supported modes in the CLI.

[github-actions]

PR Code Suggestions ✨

Category	Suggestion	Score
Possible bug	Add conditional logic to ensure preview_csv_data is only called for local datasets The conditional logic for determining whether to preview the CSV file based on the dataset type (public or local) is unclear and potentially missing. It's important to implement a clear conditional check to ensure that preview_csv_data is called only for local datasets and not for public datasets. mle/agents/advisor.py [58] -the CSV file or not if the dataset is a public dataset. +if not is_public_dataset: + preview_csv_data(dataset_path) Suggestion importance[1-10]: 8 Why: The suggestion addresses a potential bug by clarifying when to call preview_csv_data, which is crucial for correct functionality when handling different dataset types.	8
Maintainability	Refactor mode handling to use a dictionary for scalability and maintainability Consider handling additional modes in the start function or providing a more generic error handling mechanism. This would make the function more flexible and maintainable, especially if more modes are to be added in the future. mle/cli.py [52-59] -if mode == 'baseline': - # Baseline mode - return workflow.baseline(os.getcwd(), model) -elif mode == 'report': - # Report mode - return ctx.invoke(report, model=model, visualize=False) -else: - raise ValueError("Invalid mode. Supported modes: 'baseline', 'report'.") +mode_functions = { + 'baseline': lambda: workflow.baseline(os.getcwd(), model), + 'report': lambda: ctx.invoke(report, model=model, visualize=False) +} +try: + return mode_functions[mode]() +except KeyError: + raise ValueError(f"Invalid mode. Supported modes: {', '.join(mode_functions.keys())}.") Suggestion importance[1-10]: 7 Why: The suggestion improves maintainability by using a dictionary for mode handling, making the code more scalable and easier to extend with additional modes.	7
Possible issue	Verify and potentially revert the default mode to 'general' if the change was unintended The default value for the mode argument in the start command has been changed from 'general' to 'baseline'. Ensure that this change is intentional and correctly documented, as it alters the default behavior of the command. mle/cli.py [43] -@click.argument('mode', default='baseline') +@click.argument('mode', default='general') # if reverting to previous default Suggestion importance[1-10]: 5 Why: The suggestion is valid as it prompts verification of an intentional change in default behavior, which is important for maintaining expected functionality.	5