ttimbers
commented
3 years ago @hfboyce - I did not do a thorough job of checking the incorrect messages to make sure they are correct and helpful. Please check this over when you address my feedback.
Closed ttimbers closed 3 years ago
ttimbers
commented
3 years ago @hfboyce - I did not do a thorough job of checking the incorrect messages to make sure they are correct and helpful. Please check this over when you address my feedback.
hfboyce
commented
3 years ago This question asks what plot type is most appropriate to compare the distribution of two columns, technically, all answers are correct if one of the columns was categorical... We should edit this to be clear that we are talking about two numeric/quantitative columns, not categorical/qualitative data.
Updated to include numeric/quantitative and bolded.
This is almost identical to the question above, but specifies numeric, but does not have any overlapping options, which is strange. I think the point here is plotting two numerical columns where overplotting is an issue. Let's add that to the question. Also, let's remove area chart and put scatter plot in its place (since it is also good for plotting two distributions, but only if overplotting does not occur)
Ah yes. So I added overplotting here (I think that's what I was thinking here but forgot to put that in the question) and added scatter instead of area.
Change "For which scores are pokemon most saturated in the plot?" > "For which range of pokemon attack and defence scores do we have the most observations?"
The answer is "Pokemon with average defense and attack scores.", however, I might have said low defense low and attack scores... Maybe we should change this to be "Pokemon with mid-low defense and mid-low attack scores."?
Changed Question and solution answer.
Please add more scaffolding to the coding question here. It is the first one of the module, and folks likely need a reminder of all the arguments needed.
Added the arguments or the argument values now for this question
These ranges are FAR too close together. Basically, we just want them to identify the bottom left-hand part, which would be 185-200 for length, and 3200-4000 for mass. Other options we could offer would be:
- 185-200 for length, and 4600-5200 for mass
- 205 - 220 for length, and 3200-4000 for mass
- 205 - 220 for length, and 4600-5200 for mass
Updated to suggested options.
Please add more scaffolding to the coding question here. Also what is generation? You need to explain this for non-pokemon experts...
Added this:
For those less familiar with Pokemon, a Pokemon's generation is determined by the ara it was introduced. For example the first generation spans Pokemon characters that were released in video games from 1996-1999. The 7th generation are Pokemon introduced from games released between 2016-2019.
According to Wikipedia, A Pokémon's type is "an elemental attribute determining the strengths and weaknesses of each Pokémon and its moves. Pokémon take double damage from attacking moves of types they are weak to and half damage from moves they resist. These type matchups offset one another in rock–paper–scissors-style relationships."
And more scaffolding in the exercise.
- First true/false Q has no false option
Fixed!
- This question title seems a bit off...
Changed to "Exploring EDA"
- there is no correct answer here
The correct answer is "Creating a predictive model from the data" which wasn't showing up because of formatting (uuhhggg)
culmen_length_mm and flipper_length_mm distributions. " and " The culmen_length_mm and flipper_length_mm distributions both appear to be bimodal." in the popups.
- The first true/false question here is a repeat from an earlier chapter
My bad 🤦♀️
Cannot select any options from the "What species of penguins and from what island are most prominent in our data?" question
Fixed this!
Chapter 3 questions: Which plot is it
Question 1
This question asks what plot type is most appropriate to compare the distribution of two columns, technically, all answers are correct if one of the columns was categorical... We should edit this to be clear that we are talking about two numeric/quantitative columns, not categorical/qualitative data.
Question 2
This is almost identical to the question above, but specifies numeric, but does not have any overlapping options, which is strange. I think the point here is plotting two numerical columns where overplotting is an issue. Let's add that to the question. Also, let's remove area chart and put scatter plot in its place (since it is also good for plotting two distributions, but only if overplotting does not occur)
Chapter 4 questions: Graph Analysis
Change "For which scores are pokemon most saturated in the plot?" > "For which range of pokemon attack and defence scores do we have the most observations?"
The answer is "Pokemon with average defense and attack scores.", however, I might have said low defense low and attack scores... Maybe we should change this to be "Pokemon with mid-low defense and mid-low attack scores."?
Chapter 5 questions: Heatmaps with Artic Penguins
Please add more scaffolding to the coding question here. It is the first one of the module, and folks likely need a reminder of all the arguments needed.
What flipper length and mass are most common among the penguins in our data (with the given bin size)?
These ranges are FAR too close together. Basically, we just want them to identify the bottom left-hand part, which would be 185-200 for length, and 3200-4000 for mass. Other options we could offer would be:
Chapter 7 questions: Return of the Pokemon
Please add more scaffolding to the coding question here. Also what is generation? You need to explain this for non-pokemon experts...
Chapter 9 questions: True or False: Exploratory Data Analysis (EDA)
Chapter 10 questions: Which graph is appropriate?
Question 1
Chapter 11 questions: Plotting numeric columns with Penguins
Do any of the distributions look bimodal (2 bell shapes)?
Chapter 13 questions: Repeating Categorical Columns Quick Questions!
Chapter 14 questions: Plotting Repeated Categoricals with Penguins
Cannot select any options from the "What species of penguins and from what island are most prominent in our data?" question