ChatGPT output bundle is generating only one row - it should be multiple rows

[ebrodeur]

Hello, new user here.

I’ve been searching the forums and can’t find a solution to my problem, probably because I don’t know the tools well enough. Apologies in advance.

I created a Scenario with ChatGPT → Microsoft 365 Excel.

I’m using static input text for the prompt. I have created fields for the Structured Data Definition. The Excel app is “Add Worksheet Row” and the scenario successfully adds a row from the ChatGPT output.

TheChatGPT app output is only showing one output bundle (row), however the output should contain multiple rows. How do I get it to output the remaining rows of structured data?

I’ve tried Repeaters and Iterators with the same result: one row is outputted.

What am I doing wrong? Thank you for the help.

[samliew]

Welcome to the Make community!

1. Screenshots of module fields and filters

Please share screenshots of relevant module fields and filters in question? It would really help other community members to see what you’re looking at.

You can upload images here using the Upload icon in the text editor:

2. Scenario blueprint

Please export the scenario blueprint file to allow others to view the mappings and settings. At the bottom of the scenario editor, you can click on the three dots to find the Export Blueprint menu item.

^{(Note: Exporting your scenario will not include private information or keys to your connections)}

Uploading it here will look like this:

blueprint.json (12.3 KB)

3. And most importantly, Input/Output bundles

Please provide the input and output bundles of the modules by running the scenario (or get from the scenario History tab), then click the white speech bubble on the top-right of each module and select “Download input/output bundles”.

A.

Save each bundle contents in your text editor as a bundle.txt file, and upload it here into this discussion thread.

Uploading them here will look like this:

module-1-input-bundle.txt (12.3 KB)
module-1-output-bundle.txt (12.3 KB)

B.

If you are unable to upload files on this forum, alternatively you can paste the formatted bundles in this manner:

• Either add three backticks ``` before and after the code, like this:

  ```
  <sup>input/output bundle content goes here</sup>
  ```

• Or use the format code button in the editor:
  

Providing the input/output bundles will allow others to replicate what is going on in the scenario even if they do not use the external service.

Following these steps will allow others to assist you here. Thanks!

[ebrodeur]

The blueprint, input bundle and output bundle files are attached.

blueprint.json (37.1 KB)
input-bundle.txt (4.7 KB)
output-bundle.txt (231 Bytes)

Also attached is the output from the ChatGPT UI showing the expected output contains multiple rows.

chatgpt-output.txt (3.0 KB)

Thanks for your help.

[ebrodeur]

Downloading the files is probably an inconvenience. I’m posting the text inline.

Input Bundle:

[
    {
        "model": "gpt-3.5-turbo",
        "prompt": "convert all of the following text to structured columns named DATE; FORMAT; RANK; SVOD PROVIDER; PROGRAM NAME; MINUTES WATCHED (MILLIONS); WEEKLY CHANGE IN RANKING. Start at \"streaming TV\". Date column is the current date. Format column is Streaming TV or Streaming Film.",
        "rawText": "​\nView in browser\nSubscribe Now!\n \nImage\nImage\n \nTV SCHEDULES\nTV PREMIERES\nFILM RELEASES\nBOX OFFICE INSIDER\nWEEKLY DIGITAL ROUND UP\nIntroducing the new Digital Roundup Alert, now powered by Luminate Streaming Viewership (M)!\n\nThe weekly Digital Roundup is based on ratings and data as calculated and presented in Luminate Film & TV Streaming Viewership(M). This new Luminate product measures minutes watched and displays rankings for original streaming series and films by applying a linear optimization model to a combination of ACR, public and private data sources at scale, for the fastest, most accurate streaming data available.\n\nInterested in joining the Streaming Viewership(M) beta? We’re looking for industry experts to join the program. As participants you will get a first look at this amazing new product. All we ask is that you share your feedback. Click here to join the beta waitlist!\n\nHere's what's happening in digital this week:\n\nLUMINATE FILM & TV STREAMING VIEWERSHIP(M) RATINGS DATA\n\nStreaming from 3/29/2024 through 4/04/2024\n\nRANK; SVOD PROVIDER; PROGRAM NAME; MINUTES WATCHED (MILLIONS); WEEKLY CHANGE IN RANKING\n\nLuminate Top 10 Streaming TV Shows\n\n1. Netflix; 3 BODY PROBLEM; 964.1; NO CHANGE\n2. Netflix; TESTAMENT: THE STORY OF MOSES; 707.8; UP\n3. Netflix; THE GENTLEMEN; 563.4; NO CHANGE\n4. Netflix; HOMICIDE: NEW YORK; 324.3; DOWN\n5. Netflix; IS IT CAKE (SEASON 3); 303.6; NEW\n6. Netflix; BUYING BEVERLY HILLS (SEASON 2); 242.2; DOWN\n7. Apple TV+; PALM ROYALE; 198.0; DOWN\n8. Netflix; PHYSICAL: 100 (SEASON 2); 195.1; NO CHANGE\n9. Prime Video; INVINCIBLE (SEASON 2); 126.8; UP\n10. Netflix; LOVE IS BLIND (SEASON 6); 125.8; DOWN\n\nLuminate Top 10 Streaming Films\n\n1. Prime Video; ROAD HOUSE; 576.2; NO CHANGE\n2. Netflix; THE WAGES OF FEAR; 246.9; NEW\n3. Netflix; HEART OF THE HUNTER; 203.5; NEW\n4. Netflix; NO PRESSURE; 164.5; UP\n5. Netflix; THE BEAUTIFUL GAME; 125.2; NEW\n6. Prime Video; RICKY STANICKY; 121.6; DOWN\n7. Netflix; IRISH WISH; 110.9; DOWN\n8. Netflix; DAMSEL; 108.2; DOWN\n9. Netflix; SHIRLEY; 92.3; DOWN\n10. Netflix; LEAVE THE WORLD BEHIND; 78.0; NO CHANGE\n\n \n\n \n\nImage\n \nImage\nImage\nImage\n©2024 Luminate Data LLC, 5670 Wilshire Blvd,Los Angeles, CA 90036 USA. All Rights Reserved.\nPrivacy Policy | Unsubscribe | Terms of Use\n                                                           \nImage\nThis email was sent by: Luminate Film & TV\n11175 Santa Monica Boulevard Los Angeles, CA, 90025, US\n\nUpdate Profile Image",
        "parameters": [
            {
                "name": "Date",
                "type": "string",
                "isRequired": true,
                "description": "Current date",
                "valueExamples": [
                    "4/1/24"
                ]
            },
            {
                "name": "Format",
                "type": "string",
                "isRequired": true,
                "description": "Streaming TV or Streaming Film",
                "valueExamples": [
                    "Streaming TV or Streaming Film"
                ]
            },
            {
                "name": "Rank",
                "type": "number",
                "isRequired": true,
                "description": "Chronological order",
                "valueExamples": [
                    "2"
                ]
            },
            {
                "name": "Provider",
                "type": "string",
                "isRequired": true,
                "description": "SVOD Provider",
                "valueExamples": [
                    "Netflix"
                ]
            },
            {
                "name": "Program",
                "type": "string",
                "isRequired": false,
                "description": "Program Name",
                "valueExamples": [
                    "3 BODY PROBLEM"
                ]
            },
            {
                "name": "MinutesWatched",
                "type": "number",
                "isRequired": false,
                "description": "Minutes Watched (Millions)",
                "valueExamples": [
                    "964.1"
                ]
            },
            {
                "name": "WeeklyChange",
                "type": "string",
                "isRequired": true,
                "description": "Weekly Change in Ranking",
                "valueExamples": [
                    "NO CHANGE"
                ]
            }
        ]
    }
]

This is the output bundle. The fields are correct, however I’m expecting more than 1 row.

[
    {
        "Date": "4/1/24",
        "Format": "Streaming TV",
        "Rank": 1,
        "Provider": "Netflix",
        "Program": "3 BODY PROBLEM",
        "MinutesWatched": 964.1,
        "WeeklyChange": "NO CHANGE"
    }
]

This is the raw output from ChatGPT based on the text prompt.

| DATE       | FORMAT         | RANK | SVOD PROVIDER | PROGRAM NAME             | MINUTES WATCHED (MILLIONS) | WEEKLY CHANGE IN RANKING |
|------------|----------------|------|----------------|-------------------------|----------------------------|--------------------------|
| 04/08/2024 | Streaming TV   | 1    | Netflix        | 3 BODY PROBLEM          | 964.1                      | NO CHANGE                |
| 04/08/2024 | Streaming TV   | 2    | Netflix        | TESTAMENT: THE STORY OF MOSES | 707.8            | UP                       |
| 04/08/2024 | Streaming TV   | 3    | Netflix        | THE GENTLEMEN           | 563.4                      | NO CHANGE                |
| 04/08/2024 | Streaming TV   | 4    | Netflix        | HOMICIDE: NEW YORK      | 324.3                      | DOWN                     |
| 04/08/2024 | Streaming TV   | 5    | Netflix        | IS IT CAKE (SEASON 3)   | 303.6                      | NEW                      |
| 04/08/2024 | Streaming TV   | 6    | Netflix        | BUYING BEVERLY HILLS (SEASON 2) | 242.2            | DOWN                     |
| 04/08/2024 | Streaming TV   | 7    | Apple TV+      | PALM ROYALE             | 198.0                      | DOWN                     |
| 04/08/2024 | Streaming TV   | 8    | Netflix        | PHYSICAL: 100 (SEASON 2)| 195.1                      | NO CHANGE                |
| 04/08/2024 | Streaming TV   | 9    | Prime Video    | INVINCIBLE (SEASON 2)   | 126.8                      | UP                       |
| 04/08/2024 | Streaming TV   | 10   | Netflix        | LOVE IS BLIND (SEASON 6)| 125.8                      | DOWN                     |
| 04/08/2024 | Streaming Film | 1    | Prime Video    | ROAD HOUSE              | 576.2                      | NO CHANGE                |
| 04/08/2024 | Streaming Film | 2    | Netflix        | THE WAGES OF FEAR       | 246.9                      | NEW                      |
| 04/08/2024 | Streaming Film | 3    | Netflix        | HEART OF THE HUNTER     | 203.5                      | NEW                      |
| 04/08/2024 | Streaming Film | 4    | Netflix        | NO PRESSURE             | 164.5                      | UP                       |
| 04/08/2024 | Streaming Film | 5    | Netflix        | THE BEAUTIFUL GAME      | 125.2                      | NEW                      |
| 04/08/2024 | Streaming Film | 6    | Prime Video    | RICKY STANICKY          | 121.6                      | DOWN                     |
| 04/08/2024 | Streaming Film | 7    | Netflix        | IRISH WISH              | 110.9                      | DOWN                     |
| 04/08/2024 | Streaming Film | 8    | Netflix        | DAMSEL                  | 108.2                      | DOWN                     |
| 04/08/2024 | Streaming Film | 9    | Netflix        | SHIRLEY                 | 92.3                       | DOWN                     |
| 04/08/2024 | Streaming Film | 10   | Netflix        | LEAVE THE WORLD BEHIND  | 78.0                       | NO CHANGE                |

As previously mentioned, do I need to use an array, or repeater, or iterator to get more than 1 row in the output bundle?

Thanks for the help.

[samliew]

Instead of this

I think you’ll need to define the array of objects.

and you put the item definition in the “Object Definitions” section below

[ebrodeur]

Hello Samuel and thanks for the reply.

I revised the OpenAI App per your suggestion. This provided the desired Output Bundles - see image.

I added an Iterator between OpenAI and the Excel App.

This is the configuration in the Excel App.

The Excel App accepted each Output Bundle and populated the worksheet as desired. It didn’t work the first time(s) but it does now. I’m sure I tweaked something to get it going.

I cloned this Scenario and changed Data Definition 1 from “Array (objects)” to “Object.” Under that, I added a Property for each field. I did not add them under Object Definitions. See image.

This Scenario worked successfully as well for populating the rows into my Excel worksheet.

Is one of these methods “better” to use over the other?

Thank you for the help - much appreciated!

[samliew]

No problem, glad I could help!

There is no better method as long as it works and meets your requirements for your scenario.

1. If you have a new question in the future, please start a new thread. This makes it easier for others with the same problem to search for the answers to specific questions, and you are more likely to receive help since newer questions are monitored closely.

2. The Make Community guidelines encourages users to try to mark helpful replies as solutions to help keep the Community organized.

This marks the topic as solved, so that:

others can save time when catching up with the latest activity here, and

• allows others to quickly jump to the solution if they come across the same problem

To do this, simply click the checkbox at the bottom of the post that answers your question:

3. Don’t forget to like and bookmark this topic so you can get back to it easily in future!

samliew – request private consultation