C5W3_Assignment1 Wrong output while using 'st', 'nd', 'rd' and 'th' in the input dates

Just curios if anyone else has already noticed that the output of the models when entering dates in either of the following formats:

"DDst of MM YYYY",
"DDnd of MM YYYY", 
"DDrd of MM YYYY", 
"DDth of MM YYYY",

seem to be wrong:

Or am I making a mistake somewhere?


Hello @mrgransky,

If you pass all the tests and the autograder, then you can be pretty confident that you didn’t make major mistakes.

However, not even “st” works, as shown in the examples you shared.

My suggestion for you is to look at the training data, and see how well date formats with “th”, “st”, “rd” and “nd” are represented. I did it and I was pretty surprised. However, I will leave it to you to do the checking yourself. :wink:

When you check, write some simple Python code to keep only relevant entries, and inspect them carefully.

Btw, it is a very good observation.

Good luck.