dgrtwo / data-screencasts Goto Github PK

in line ... show_metrics <- create_metrics(show_summarized)

it throws an error from my part. I am running the same code you shared.
The error is:-
Error in stringr::str_split(y$name, "_")[[1]] : subscript out of bounds

Dear David,
There was a typo in the code. That's why the error with linear-reg and XGBoost was high on the Kaggle SLICED submission.
write_csv(attempt1, "~/Desktop/attempt3.csv")

It should have been.
write_csv(attempt3, "~/Desktop/attempt3.csv")

Could you please add a license to this repository?

Just loving the screencasts you have been doing

On the penguin code. I'm not sure if you want lines 924 onwards to be in your final version

I think the first few of these lines were overtaken by further analyses. If you want them in, you may want to include the code that creates the 'model' object: otherwise an error occurs

model <- logistic_reg(mode= "classification") %>% 
  set_engine("glm") %>% 
  fit(species ~ bill_length_mm, data= training_data)

From my recollection, the final part of the rmd is just you testing a vignette to understand a function and, though useful in your thought process and as a good tip, is not directly relevant to the penguin model

David,
Thanks, Another great screencast today.
Please push today's code up.

TX,
David

Hi, I really appreciate your screencast and have learned a lot from it. While studying your nyc-restaurants screencast code. I found out a place where the code breaks. It worked when you were doing the screencast, but it doesn't work at the moment, even after I tried to tweak it for a number of times. Hopefully, you can take a look and make some changes accordingly.

The first one is
cuisine_conf_ints <- by_dba %>% add_count(cuisine) %>% filter(n > 100) %>% nest(-cuisine) %>% mutate(model = map(data, ~ t.test(.$avg_score))) %>% unnest(map(model, tidy))
which throws me an error message Error: object 'model' not found

Thanks!