How to get chatGPT to recognize tables in a prompt

Can someone let me know how to get chatGPT to recognize a table in a prompt without actually specifying that the following is a table. For example

If I wrote:

Please summarize the following text:
[By default, Auto Loader schema inference seeks to avoid schema evolution issues due to type mismatches. For formats that don’t encode data types (JSON and CSV), Auto Loader infers all columns as strings (including nested fields in JSON files).

File format Default inferred data type
JSON String
CSV String
Avro Types encoded in Avro schema
Parquet Types encoded in Parquet schema

he Apache Spark DataFrameReader uses different behavior for schema inference.]

As you see from the above, I have enclosed the text with square brackets . However, within the text there is the following table.

File format Default inferred data type

JSON String
CSV String
Avro Types encoded in Avro schema
Parquet Types encoded in Parquet schema

Will chatGPT recognize the above as a table? Or will need to do the following for chatGPT to recognize as a table

File format Default inferred data type
JSON String
CSV String
Avro Types encoded in Avro schema
Parquet Types encoded in Parquet schema

Thanks

1 Like

Experimentation suggests it will likely recognize it as tabular data, but may not include the formatted table in the output unless prompted. After the bracketed text, include a directive such as “Include a table in the output.” or “Include tables in the output.”

Additionally, ChatGPT may not separate the fields as you want (possibly resulting in a different number of columns than you expect) without something to differentiate the columns. Make sure you separate the columns with fairly standard markers, such as tab characters, or pipes, or some combination (e.g. " | ", tab-pipe-tab). If you want a widely used convention with some tool support, try a Markdown table (note: tables are not yet a part of the CommonMark standard, but are available as extensions in a number of implementations).