Python+#

Overview#

Finplicity provides basic operations for cleaning and reformatting tables in Excel. However, when you need more power and flexibility, Python+ offers a powerful solution.

Python+ combines the capabilities of Python and AI without requiring you to write Python code yourself. It allows you to manipulate data across multiple tables simultaneously. Simply describe your task in natural language, and Python+ uses external AI tools like Gemini to generate Python scripts for you. It’s beginner-friendly and a great tool for learning Python through real-world use.

Key features:

  • AI-powered code generation: Just type your goal using natural language and click the “Ask AI” button. The corresponding Python code will be generated automatically.

  • Inline previews: See the output results before writing them back into Excel worksheets—making it easier to organize and verify your data layout.

  • View generated scripts: Curious about the code behind the scenes? You can view the AI-generated Python scripts to better understand how the solution works and learn from it.

Videos#

Examples#

Example 1: Count Employees by Department#

  1. Open the Python+ tab in the ribbon, and we have the following data.

  2. Add a new table:

    • Name: employees

    • Location: A1:D6

  3. Compose Prompt:


Load the `employees` table and print a count of employees by department.

  1. Click Show Full Prompt

  2. Click Ask AI

  3. Click Preview Output to see the department counts.

  4. Click Show Scripts to inspect the generated Python code.

  5. Click Select Output Location to place the results in your sheet.

Example 2: Compute Monthly Revenue#

  1. Open the Python+ tab in the ribbon, and we have the following data.

  2. Add a new table:

  • Name: sales_data

  • Location: A1:D6

  1. Compose Prompt:


From `sales_data`, filter for orders in 2024, then compute monthly revenue totals.

  1. Click Show Full Prompt

  2. Click Ask AI

  3. Click Preview Output to verify the monthly totals.

  4. Click Show Scripts (The script should parse dates and group by month.)

  5. Click Select Output Location to insert the revenue table.

Example 3: Train & Evaluate a Model#

  1. Open the Python+ tab in the ribbon, and we have the following data.

  2. Add a new table:

  • Name: iris

  • Location: A1:E6

  1. Compose Prompt:


Load the `iris` table, split into train/test (80/20), train a scikit-learn decision tree to predict species, and print the accuracy score.

  1. Click Show Full Prompt

  2. Click Ask AI

  3. Click Preview Output to view the accuracy.

  4. Click Show Scripts (Look for train_test_split, DecisionTreeClassifier, and clf.score.)

  5. Click Select Output Location to place the result in your workbook.

FAQ#

  • Will my data be exposed to AI tools?

No, your data is not shared with external AI tools. We generated some sample (fake) data before sending the prompt to AI. You can review exactly what’s being sent by clicking “Show full prompt” button.

  • Why I got error “No Python code found in the AI response.”?

You may encounter this error when clicking the “Ask AI” button. This usually means either we didn’t receive a response from the AI, or the response didn’t contain any Python code.

Here are two things you can try to resolve the issue:

  1. Change the AI vendor or model in the AI settings. A different model may produce better results.

  2. Refine your prompt to more clearly describe your goals. This helps the AI better understand what you’re trying to achieve.