What are the unexpected effects of the combination of ChatGPT and form plug-ins?
Abstract: This article was originally published on CSDN by the technical team of Grape City. Please indicate the source of the reprint: Grape City official website , Grape City provides developers with professional development tools, solutions and services to empower developers.
ChatGPT has been popular for a long time. I want to explore the usage scenarios of ChatGPT in the table. After thinking for a long time, I sorted out three points: 1.
Use assistant: provide operation guides according to requirements, and generate formulas according to descriptions.
2. Data analysis: Provide data analysis suggestions for the data in the table, and analyze the data according to the description.
3. Work assistant: combine SpreadJS data content to generate analysis reports, emails, and make predictions.
After sorting out the case and writing the demo, I didn't think why I didn't ask ChatGPT directly. He told me: I
have to sigh, the writing is awesome, more comprehensive than I thought. But before that, everyone must be curious, what exactly is SpreadJS? See how ChatGPT answered:
That's right, SpreadJS is a JavaScript-based spreadsheet control that provides an Excel-like spreadsheet experience by displaying data in tabular form, a high-speed calculation engine, and hundreds of statistical and financial functions and formulas. It is easy to implement, scalable and flexible. In particular, it can help enhance your application and transform content from a simple set of data into a more useful, understandable Excel-like dashboard.
Well, let’s get down to business, let’s take a look at what kind of sparks will be created when SpreadJS meets ChatGPT.
1. ChatGPT formula
-
**Filter data:**By using the GPT.FILTER formula, data filtering can be conveniently performed. Just select the data, add a description, and GPT.FILTER will return a dynamic array filled to the cell.
2. Generate a description
Use the GPT.QUERY formula to ask GPT to get the result of the question. In the question, the cell data in the table can be referenced, and description information related to the description can be generated according to the cell content.
2. Formula generation, analysis, optimization
1. Formula generation
Through the use of natural language description, an executable Excel formula is generated and inserted into the cell after checking. Very friendly to spreadsheet beginners.
2. Formula analysis
Complicated nested formulas in cells will return descriptions after GPT analysis to help users quickly understand the meaning of the formulas in the table
3. Formula optimization
For complex formulas, multiple optimization schemes can be given to reduce the length of the formula or improve the efficiency of the formula operation.
3. Data Analysis
1. Pivot Table Suggestions
Select the tabular data containing the title, give suggestions for pivot table analysis, including how to set rows, columns, and arrays, and explain the meaning of the analyzed data.
2. Data analysis
For the selected regional data, the analysis results can be obtained directly according to the description, eliminating the cumbersome screening and sorting process. You can also ask ChatGPT to give a specific analysis process for reference.
4. Report and email content generation
In addition to simple question replies, ChatGPT can conduct dialogues, modify question replies based on contextual content, or obtain further analysis. In the example below, ChatGPT gives the data analysis results of multiple dimensions, we choose one of the options, and asked him to generate an analysis email to Ted.
Click here to watch the demo video
The ChatGPT plugin provides an intelligent natural language interaction experience for SpreadJS users. Users can use natural language queries and instructions to interact with spreadsheets, without having to manually write complex formulas and configure functions such as data penetration; it is faster and easier to parse and execute user data operation requests through ChatGPT, which improves work efficiency; at the same time, ChatGPT Some analysis suggestions and error detection can also be given, improving the quality of the spreadsheet content.
Of course, there are still two obvious problems with the ChatGPT plugin:
- Limitations of model predictions, ChatGPT is an opportunistic language model, and its performance is limited by model training and data sets. For complex problems, the results may not be returned accurately, and the returned result format may not be analyzed by the plug-in.
- Relying on external services, the use of the plug-in needs to call the ChatGPT service, and a reliable network connection must be ensured when using it. At the same time, the table data in SpreadJS must be sent to the ChatGPT server, and data security and privacy are also issues to be considered.
Overall, the ChatGPT plug-in can provide users with intelligent natural language interaction and data manipulation capabilities, but it also needs to pay attention to the limitations of data security and language models. Looking forward to more powerful AI tools for data analysis.
If you are interested in SpreadJS, you can also visit the official website of Putaocheng to learn about the related features of SpreadJS.
Extension link:
Teach you how to design a large screen for visual data
Low-code combined with report plug-ins to build a complex reporting system