TA comments

[ChadFibke]

Coding Style

The good :thumbsup:

• Simple effect code!

• Good name choice for some of your created objects! They are informative!

Needs work :thumbsdown:

• There are still many nested objects with uninformative names. p1,p2,p3 wont help you if you are trying to trouble shoot issues when using plot_grid.

• Try to break up some of your ggplot code for better readability:

  
  gapminder %>% 
  ggplot( aes( gdpPercap,lifeExp ) ) + 
  scale_x_log10() +
  geom_point(size = 0.1, alpha = 0.5) + 
  facet_wrap( ~ continent )+
  theme_bw() +
  geom_smooth( method = "lm", size = 0.3, color = "red" )

* Some spaces in your code are effective while others are not. Please see [here](http://style.tidyverse.org/) for more consistent spacing.
# Coding Strategy

### The good :thumbsup:

* Broke the challenges in the manageable bite size pieces by using consistent markdown syntax.

* Good amount of in line comments explaining the output. This also helps to guide the analysis.

### Needs work :thumbsdown:
* NA

# Graphs

### The good :thumbsup:

* Good incorporation of challenge questions/tasks into your general home work! 

* Effect use of reorder to get better ordered graphs!

### Needs work :thumbsdown:

* Your grobs are overlapping in your arranged grid!

* For next time: try playing around with theme functions. Theme can really help manipulate anything from the position, size, color, and even the font style of text, legends and more! For a larger scope, please look [here](https://ggplot2.tidyverse.org/reference/theme.html). Also, please insure that all your graphs have titles, and that all axes have been revamped to be publication quality (no under scores and use of capitals)! 

* Some of your more complex graphs are smooshed together! Try reformatting the rmd file code chunks the following to get bigger output:
"{r, fig.width = nigger_num, fig.height = bigger_num}"

* Please use color responsibly! Having too many groups can actually make color work against you! This also goes for plot annotation! All that extra work you did to annotate the lines are now just balled up in the corner of your graph :frowning:

# Tables 

### The good :thumbsup:

* Tables look great! Thanks for exploring the grid extra package, and playing around with some table grobs! 

### Needs work :thumbsdown:

* NA

# Creativity & Ease of Access

### The good :thumbsup:

* Homework was easy to get to and read!

* Went above and beyond with exploration of many packages and concepts!

### Needs work :thumbsdown:

** You have several unnecessary files (DS store, README.html, and hw03.html), which will overwhelm readers as the repo grows! You can keep these files, but please store them in your supplementary files folder.

* Good try this your toc, but it did not work. I checked your YAML header in your rmd, and you did not specify a toc section there.

[rasiimwe]

Hi @ChadFibke

Thanks for your awesome comments. I will apply the advice in my following assignments. Some areas are still not clear.

When you said "There are still many nested objects with uninformative names. p1,p2,p3 won't help you if you are trying to troubleshoot issues when using plot_grid." Did you mean I should use more elaborative names in the example below 👇 ? Say instead of p1 I should use plot_1?

p1 <- mean_lifeExp_cont %>% 
  ggplot(aes(year, mean_lifeExp,color=continent))+
  geom_line(size=0.5)+
  scale_color_manual(values=continent_colors)+ #I am using colors provided by the gapminder package
  geom_point(shape=3,stat='summary', fun.y=sum) +
  stat_summary(fun.y=sum, geom="line")+ 
  theme_bw()+
  scale_x_continuous(breaks=mean_lifeExp_cont$year) 

p2<- mean_lifeExp_cont %>%
  ggplot(aes(year, continent,  fill = mean_lifeExp)) +
  geom_tile(color = "gray40") +
  scale_x_continuous(expand=c(0,0)) +
  scale_fill_gradient(low = "red", high = "midnightblue")+
  theme_classic()   

p3<-mean_lifeExp_cont %>% 
  ggplot(aes(reorder(continent, mean_lifeExp), mean_lifeExp))+
  geom_col(aes(fill = mean_lifeExp), size=0.1)+ 
  coord_flip()+#flip the axes for a horizontal representation - I prefered horizontal bars
 scale_fill_gradient2(low = "white", mid="red", high = "midnightblue")+theme_light()+labs(y="Mean Life   Expectancy", x="Continent") #because of coord flip, the labels change too!

plot_grid(p1, p2, p3, labels = "AUTO", nrow=3)

I removed y-axis labels because I thought it redundant to have the y axis labels repeated on plots that appear side by side - it may look like the plots are overlapping - I removed labels as seen in my code 😄 .

I also tried to show other representations of selecting outliers and elaborated on why it's not helpful to plot the way I did in option 2 of section4. Maybe I should avoid walking my readers through other variations out there and what to avoid. I think its a lot of work for me 😄 as well yet not all visitors will take the time to read everything. Thanks for pointing that out.

Thanks for the helpful comments. Thanks for the helpful links as well. Many thanks

[ChadFibke]

Hey,

Yes you should use more specific names even when you are using the plots to plot in a more complex grid!

You are right it would be redundant to keep both y axis; however, I'm talking about the very first table plot combo. All of the P's look like they are slightly cut off. Your heat maps where fine!

There was a lot to unpack in your assignment, and I missed that. I look at it again and yes you are right, but your Europe line graph is still a bit messy when it comes to labels and colors! For next time can you try to only label/color country lines with values higher than X (mission impossible), and even explaining your output, and reasoning with point form?

Other than that you did really good!

[rasiimwe]

@ChadFibke

Thanks for the clarity. I will do as advised.

Many thanks Rebecca