chatgpt: Create fake data for your application and skip learning a library to do t
One very common task i have is to create fake data to test an application, for example we might need a csv file with name, last name , age…
chatgpt: Create test data for your application and skip learning a library to do this!
One very common task i have is to create test data to test an application, for example we might need a csv file with name, last name , age and other data, There are many libraries to do this but lets see how we can do the same task using chatgpt, the main advantage of using chatgpt for such tasks is that you eliminate the need to learn a new library which can be time consuming.
In order to perform the test you will need to create a Python script that will query chatgpt or to ask chatgpt from the web site, but you can also try a command like tool i have create for such tasks
kpatronas/shellgpt: Query chatgpt from your shell (github.com)
Its a very handy shell script that with minimal configuration you can ask chatgpt from the command line prompt like this.
$ chatgpt -p "Tell me a joke"
Q: What did the fish say when it hit the wall?
A: Dam!How we can generate test data with chatgpt
When dealing with chatgpt queries you must be very precice on what you ask because the same question can vary from run to run
For example i asked for “Create a table with 5 rows having headers name,last-name,age,gender with random values”, the result is correct but it added the “|” character as a separator, which is not a mistake since we didn’t mention the separator.
chatgpt -p "Create a table with 5 having headers name,last-name,age,gender with random values"
Name | Last-Name | Age | Gender
-----|-----------|-----|-------
John | Smith | 25 | Male
Jane | Doe | 32 | Female
Bob | Johnson | 28 | Male
Sara | Miller | 22 | Female
Alex | Williams | 30 | MaleBeing more precise gives consistent results
chatgpt -p "Create a table with 20 having headers name,last-name,age,gender with random values, comma separated."
Name,Last-Name,Age,Gender
John,Smith,25,Male
Jane,Doe,30,Female
Michael,Williams,45,Male
Emily,Jones,35,Female
Jacob,Brown,20,Male
Sophia,Davis,27,Female
Joshua,Miller,40,Male
Emma,Wilson,32,Female
Matthew,Taylor,22,Male
Olivia,Anderson,24,Female
Daniel,Thomas,50,Male
Ava,Jackson,28,Female
Christopher,White,38,Male
Isabella,Harris,31,Female
Andrew,Martin,23,Female
Mia,Thompson,26,Female
Anthony,Garcia,41,Male
Abigail,Martinez,29,Female
David,Robinson,48,Male
Elizabeth,Clark,33,FemaleWe can even fine tune parameters for example the number of males vs females
$ chatgpt -p "Create a table with 20 rows having headers name,last-name,age,gender with random values, comma separated. with 15 males and 5 females"
Name,Last-Name,Age,Gender
John,Smith,25,Male
James,Johnson,30,Male
Michael,Williams,35,Male
David,Jones,40,Male
Robert,Brown,45,Male
Mark,Davis,50,Male
Steven,Miller,55,Male
Joseph,Wilson,60,Male
Charles,Moore,65,Male
Christopher,Taylor,70,Male
Daniel,Anderson,75,Male
Paul,Thomas,80,Male
Richard,Jackson,85,Male
Brian,White,90,Male
Kevin,Harris,95,Male
Jason,Martin,20,Male
Jeffrey,Thompson,25,Male
Anthony,Garcia,30,Male
Mary,Martinez,35,Female
Patricia,Robinson,40,Female
Linda,Clark,45,Female
Barbara,Rodriguez,50,Female
Jennifer,Lewis,55,Femaleconclusion
chatgpt might probably will not replace developers and engineers of any kind any time soon, but its a valuable tool that can help us to accomplish tasks that will take a lot of time by learning a library.
