Geography 13

# Geography 13
## Lecture 08: Working with Tables Wrap Up
### Mike Johnson

---

# Picking back up!

We’ve covered many topics on how to manipulate and reshape a single data.frame:

**Last Week**: Data type and data structures

**Monday**: `data.frame` manipulation

**Tuesday**: `data.frame` visualization

**Wednesday**: Joining and pivoting `data.frames`

---

# Yesterdays Assignment

**Question 1**: Make a _faceted_ plot of the cumulative cases & deaths by USA region. Your _x_ axis should be the _date_ and the _y_ axis value/count. To do this you will need to join and pivot the COVID-19 data.

```r
library(tidyverse)
url = 'https://raw.githubusercontent.com/nytimes/covid-19-data/master/us-counties.csv'
covid = read_csv(url)
head(covid)
```

```
# A tibble: 6 x 6
  date       county    state     fips  cases deaths
  <date>     <chr>     <chr>     <chr> <dbl>  <dbl>
1 2020-01-21 Snohomish Washingt… 53061     1      0
2 2020-01-22 Snohomish Washingt… 53061     1      0
3 2020-01-23 Snohomish Washingt… 53061     1      0
4 2020-01-24 Cook      Illinois  17031     1      0
5 2020-01-24 Snohomish Washingt… 53061     1      0
6 2020-01-25 Orange    Californ… 06059     1      0
```
]

```r
region = data.frame(state = state.name, region = state.region)
head(region)
```

```
       state region
1    Alabama  South
2     Alaska   West
3    Arizona   West
4   Arkansas  South
5 California   West
6   Colorado   West
```
]

---

count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
*inner_join(covid, region, by = "state")
```
]
 
.panel2-q11-auto[

```
# A tibble: 1,451,491 x 7
   date       county    state    fips  cases deaths
   <date>     <chr>     <chr>    <chr> <dbl>  <dbl>
 1 2020-01-21 Snohomish Washing… 53061     1      0
 2 2020-01-22 Snohomish Washing… 53061     1      0
 3 2020-01-23 Snohomish Washing… 53061     1      0
 4 2020-01-24 Cook      Illinois 17031     1      0
 5 2020-01-24 Snohomish Washing… 53061     1      0
 6 2020-01-25 Orange    Califor… 06059     1      0
 7 2020-01-25 Cook      Illinois 17031     1      0
 8 2020-01-25 Snohomish Washing… 53061     1      0
 9 2020-01-26 Maricopa  Arizona  04013     1      0
10 2020-01-26 Los Ange… Califor… 06037     1      0
# … with 1,451,481 more rows, and 1 more variable:
#   region <fct>
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
inner_join(covid, region, by = "state") %>%
* count(region)
```
]
 
.panel2-q11-auto[

```
# A tibble: 4 x 2
  region             n
  <fct>          <int>
1 Northeast     104137
2 South         661480
3 North Central 482878
4 West          202996
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
inner_join(covid, region, by = "state") %>%
  count(region) %>%
* mutate(tot = sum(n))
```
]
 
.panel2-q11-auto[

```
# A tibble: 4 x 3
  region             n     tot
  <fct>          <int>   <int>
1 Northeast     104137 1451491
2 South         661480 1451491
3 North Central 482878 1451491
4 West          202996 1451491
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
inner_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

*full_join(covid, region, by = "state")
```
]
 
.panel2-q11-auto[

```
# A tibble: 4 x 3
  region             n     tot
  <fct>          <int>   <int>
1 Northeast     104137 1451491
2 South         661480 1451491
3 North Central 482878 1451491
4 West          202996 1451491
```

```
# A tibble: 1,488,568 x 7
   date       county    state    fips  cases deaths
   <date>     <chr>     <chr>    <chr> <dbl>  <dbl>
 1 2020-01-21 Snohomish Washing… 53061     1      0
 2 2020-01-22 Snohomish Washing… 53061     1      0
 3 2020-01-23 Snohomish Washing… 53061     1      0
 4 2020-01-24 Cook      Illinois 17031     1      0
 5 2020-01-24 Snohomish Washing… 53061     1      0
 6 2020-01-25 Orange    Califor… 06059     1      0
 7 2020-01-25 Cook      Illinois 17031     1      0
 8 2020-01-25 Snohomish Washing… 53061     1      0
 9 2020-01-26 Maricopa  Arizona  04013     1      0
10 2020-01-26 Los Ange… Califor… 06037     1      0
# … with 1,488,558 more rows, and 1 more variable:
#   region <fct>
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
inner_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

full_join(covid, region, by = "state") %>%
* count(region)
```
]
 
.panel2-q11-auto[

```
# A tibble: 5 x 2
  region             n
  <fct>          <int>
1 Northeast     104137
2 South         661480
3 North Central 482878
4 West          202996
5 <NA>           37077
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q11-auto[

```r
inner_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

full_join(covid, region, by = "state") %>%
  count(region) %>%
* mutate(tot = sum(n))
```
]
 
.panel2-q11-auto[

```
# A tibble: 5 x 3
  region             n     tot
  <fct>          <int>   <int>
1 Northeast     104137 1488568
2 South         661480 1488568
3 North Central 482878 1488568
4 West          202996 1488568
5 <NA>           37077 1488568
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
*left_join(covid, region, by = "state")
```
]
 
.panel2-q12-auto[

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
left_join(covid, region, by = "state") %>%
* count(region)
```
]
 
.panel2-q12-auto[

```
# A tibble: 5 x 2
  region             n
  <fct>          <int>
1 Northeast     104137
2 South         661480
3 North Central 482878
4 West          202996
5 <NA>           37077
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
left_join(covid, region, by = "state") %>%
  count(region) %>%
* mutate(tot = sum(n))
```
]
 
.panel2-q12-auto[

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
left_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

*right_join(covid, region, by = "state")
```
]
 
.panel2-q12-auto[

```
# A tibble: 5 x 3
  region             n     tot
  <fct>          <int>   <int>
1 Northeast     104137 1488568
2 South         661480 1488568
3 North Central 482878 1488568
4 West          202996 1488568
5 <NA>           37077 1488568
```

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
left_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

right_join(covid, region, by = "state") %>%
* count(region)
```
]
 
.panel2-q12-auto[

```
# A tibble: 4 x 2
  region             n
  <fct>          <int>
1 Northeast     104137
2 South         661480
3 North Central 482878
4 West          202996
```
]

---
count: false
 
##Question 1: Joins (the right tool for the job)
.panel1-q12-auto[

```r
left_join(covid, region, by = "state") %>%
  count(region) %>%
  mutate(tot = sum(n))

right_join(covid, region, by = "state") %>%
  count(region) %>%
* mutate(tot = sum(n))
```
]
 
.panel2-q12-auto[

---
# Whats causing the NAs?

- We know there are non-matching keys between COVID and region
- We know that the "extras" are in the covid data.frame (NAs produced in `left_join(covid, region)`)

```r
covid %>% 
  filter(!state %in% state.name) %>% 
  filter(date == max(date)) %>% 
  count(state)
```

```
# A tibble: 5 x 2
  state                        n
  <chr>                    <int>
1 District of Columbia         1
2 Guam                         1
3 Northern Mariana Islands     2
4 Puerto Rico                 79
5 Virgin Islands               3
```

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
*covid
```
]
 
.panel2-q13-auto[

```
# A tibble: 1,488,568 x 6
   date       county    state    fips  cases deaths
   <date>     <chr>     <chr>    <chr> <dbl>  <dbl>
 1 2020-01-21 Snohomish Washing… 53061     1      0
 2 2020-01-22 Snohomish Washing… 53061     1      0
 3 2020-01-23 Snohomish Washing… 53061     1      0
 4 2020-01-24 Cook      Illinois 17031     1      0
 5 2020-01-24 Snohomish Washing… 53061     1      0
 6 2020-01-25 Orange    Califor… 06059     1      0
 7 2020-01-25 Cook      Illinois 17031     1      0
 8 2020-01-25 Snohomish Washing… 53061     1      0
 9 2020-01-26 Maricopa  Arizona  04013     1      0
10 2020-01-26 Los Ange… Califor… 06037     1      0
# … with 1,488,558 more rows
```
]

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
covid %>%
* right_join(region, by = "state")
```
]
 
.panel2-q13-auto[

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
covid %>%
  right_join(region, by = "state") %>%
* group_by(region, date)
```
]
 
.panel2-q13-auto[

```
# A tibble: 1,451,491 x 7
# Groups:   region, date [2,092]
   date       county    state    fips  cases deaths
   <date>     <chr>     <chr>    <chr> <dbl>  <dbl>
 1 2020-01-21 Snohomish Washing… 53061     1      0
 2 2020-01-22 Snohomish Washing… 53061     1      0
 3 2020-01-23 Snohomish Washing… 53061     1      0
 4 2020-01-24 Cook      Illinois 17031     1      0
 5 2020-01-24 Snohomish Washing… 53061     1      0
 6 2020-01-25 Orange    Califor… 06059     1      0
 7 2020-01-25 Cook      Illinois 17031     1      0
 8 2020-01-25 Snohomish Washing… 53061     1      0
 9 2020-01-26 Maricopa  Arizona  04013     1      0
10 2020-01-26 Los Ange… Califor… 06037     1      0
# … with 1,451,481 more rows, and 1 more variable:
#   region <fct>
```
]

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
covid %>%
  right_join(region, by = "state") %>%
  group_by(region, date) %>%
* summarize(cases  = sum(cases),
*           deaths = sum(deaths))
```
]
 
.panel2-q13-auto[

```
# A tibble: 2,092 x 4
# Groups:   region [4]
   region    date       cases deaths
   <fct>     <date>     <dbl>  <dbl>
 1 Northeast 2020-02-01     1      0
 2 Northeast 2020-02-02     1      0
 3 Northeast 2020-02-03     1      0
 4 Northeast 2020-02-04     1      0
 5 Northeast 2020-02-05     1      0
 6 Northeast 2020-02-06     1      0
 7 Northeast 2020-02-07     1      0
 8 Northeast 2020-02-08     1      0
 9 Northeast 2020-02-09     1      0
10 Northeast 2020-02-10     1      0
# … with 2,082 more rows
```
]

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
covid %>%
  right_join(region, by = "state") %>%
  group_by(region, date) %>%
  summarize(cases  = sum(cases),
            deaths = sum(deaths)) %>%
* pivot_longer(cols = c('cases', 'deaths'))
```
]
 
.panel2-q13-auto[

```
# A tibble: 4,184 x 4
# Groups:   region [4]
   region    date       name   value
   <fct>     <date>     <chr>  <dbl>
 1 Northeast 2020-02-01 cases      1
 2 Northeast 2020-02-01 deaths     0
 3 Northeast 2020-02-02 cases      1
 4 Northeast 2020-02-02 deaths     0
 5 Northeast 2020-02-03 cases      1
 6 Northeast 2020-02-03 deaths     0
 7 Northeast 2020-02-04 cases      1
 8 Northeast 2020-02-04 deaths     0
 9 Northeast 2020-02-05 cases      1
10 Northeast 2020-02-05 deaths     0
# … with 4,174 more rows
```
]

---
count: false
 
##Question 1: Refine and Pivot (the right data for the task)
.panel1-q13-auto[

```r
covid %>%
  right_join(region, by = "state") %>%
  group_by(region, date) %>%
  summarize(cases  = sum(cases),
            deaths = sum(deaths)) %>%
  pivot_longer(cols = c('cases', 'deaths')) ->
* covid_region
```
]
 
.panel2-q13-auto[

]

---

```r
*ggplot(covid_region, aes(x = date, y = value))
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_01_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
* geom_line(aes(col = region))
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_02_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
* facet_grid(name~region, scale = "free_y")
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_03_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
* theme_linedraw()
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_04_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
* theme(legend.position = "bottom")
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_05_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
  theme(legend.position = "bottom") +
* theme(legend.position = "NA")
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_06_output-1.png" width="432" />
]

---
count: false
 
##Question 1: Plotting (table to visual!)
.panel1-q14-auto[

```r
ggplot(covid_region, aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
  theme(legend.position = "bottom") +
  theme(legend.position = "NA") +
* labs(title = "Cummulative Cases and Deaths: Region",
*      y = "Daily Cumulative Count",
*      x = "Date",
*      caption = "Daily Exercise 07",
*      subtitle = "COVID-19 Data: NY-Times" )
```
]
 
.panel2-q14-auto[
<img src="lecture-08_files/figure-html/q14_auto_07_output-1.png" width="432" />
]

---
class: center, middle, inverse
# Today: 
Wrapping up our "working with tables" unit
---

# Re-cap: Data Types

- **Values** have a type

- Values can be **assigned** to **names**

- Assignment creates a permanent **object** (within the working session)

- **Objects** can be seen in the Environment tab

- names can be referenced in scripts, Rmds, and code

- Objects have a **type** and a **class**

```r
x = 10
class(x)
```

```
[1] "numeric"
```

```r
typeof(x)
```

```
[1] "double"
```

---

# S3 Atomic Vectors

- One of the most important vector attributes is `class`, which underlies the S3 object system.

- A class attribute turns an object into an S3 object, which means it will behave differently from a regular vector when passed to a generic function.

- Every S3 object is built on top of a base type, and can store additional information in other `attributes`.

---

# S3: Use cases

1. Categorical data, where values come from a fixed set of levels recorded in factor vectors.

2. Dates (daily resolution), which are recorded in Date vectors.

3. Date-times (with second or sub-second resolution), which are stored in POSIXct vectors.

---

# Categorical

- Much like a categorical raster (think landcover), `factors` store categorical values

- A factor vector can only contain only predefined values.

- Factors are built on the `integer` data type with two attributes:

-  a `class` (factor) which makes it behave differently from integer vectors,

-  a `levels`, which defines the allowed values.

---
count: false
 
#Factors
.panel1-fac-auto[

```r
*(x = factor(c("Jan", "Feb", "Mar", "Apr")))
```
]
 
.panel2-fac-auto[

```
[1] Jan Feb Mar Apr
Levels: Apr Feb Jan Mar
```
]

---
count: false
 
#Factors
.panel1-fac-auto[

```r
(x = factor(c("Jan", "Feb", "Mar", "Apr")))
*class(x)
```
]
 
.panel2-fac-auto[

```
[1] Jan Feb Mar Apr
Levels: Apr Feb Jan Mar
```

```
[1] "factor"
```
]

---
count: false
 
#Factors
.panel1-fac-auto[

```r
(x = factor(c("Jan", "Feb", "Mar", "Apr")))
class(x)
*typeof(x)
```
]
 
.panel2-fac-auto[

```
[1] Jan Feb Mar Apr
Levels: Apr Feb Jan Mar
```

```
[1] "factor"
```

```
[1] "integer"
```
]

---
count: false
 
#Factors
.panel1-fac-auto[

```r
(x = factor(c("Jan", "Feb", "Mar", "Apr")))
class(x)
typeof(x)
*attributes(x)
```
]
 
.panel2-fac-auto[

```
[1] Jan Feb Mar Apr
Levels: Apr Feb Jan Mar
```

```
[1] "factor"
```

```
[1] "integer"
```

```
$levels
[1] "Apr" "Feb" "Jan" "Mar"

$class
[1] "factor"
```
]

---
count: false
 
#Factors
.panel1-fac-auto[

```r
(x = factor(c("Jan", "Feb", "Mar", "Apr")))
class(x)
typeof(x)
attributes(x)
*unclass(x)
```
]
 
.panel2-fac-auto[

```
[1] Jan Feb Mar Apr
Levels: Apr Feb Jan Mar
```

```
[1] "factor"
```

```
[1] "integer"
```

```
$levels
[1] "Apr" "Feb" "Jan" "Mar"

$class
[1] "factor"
```

```
[1] 3 2 4 1
attr(,"levels")
[1] "Apr" "Feb" "Jan" "Mar"
```
]

---

# Factors in the "wild"

```r
head(gapminder,1)
```

```
# A tibble: 1 x 6
  country     continent  year
  <fct>       <fct>     <int>
1 Afghanistan Asia       1952
# … with 3 more variables:
#   lifeExp <dbl>, pop <int>,
#   gdpPercap <dbl>
```

```r
head(covid_region,1)
```

```
# A tibble: 1 x 4
# Groups:   region [1]
  region  date       name  value
  <fct>   <date>     <chr> <dbl>
1 Northe… 2020-02-01 cases     1
```

```r
levels(covid_region$region)
```

```
[1] "Northeast"    
[2] "South"        
[3] "North Central"
[4] "West"         
```

```r
nlevels(covid_region$region)
```

```
[1] 4
```
---

# So what?

- Factors are stored as an integer vector with a corresponding set of character values to use when the factor is displayed.

- The ordering of factor drives some of our graphics and other functions that rely on order 
--

```r
levels(covid_region$region)
```

```
[1] "Northeast"    
[2] "South"        
[3] "North Central"
[4] "West"         
```
--

<img src="lecture-08_files/figure-html/unnamed-chunk-9-1.png" width="432" />
---

# Levels

```r
## default order is alphabetical
covid_region$region %>%
  levels()
```

```
[1] "Northeast"    
[2] "South"        
[3] "North Central"
[4] "West"         
```

```r
## order by frequency
covid_region$region %>% 
  fct_infreq() %>%
  levels()
```

```
[1] "West"         
[2] "North Central"
[3] "Northeast"    
[4] "South"        
```

---

# Levels

```r
## backwards!
covid_region$region %>% 
  fct_rev() %>%
  levels()
```

```
[1] "West"         
[2] "North Central"
[3] "South"        
[4] "Northeast"    
```

---

# Some toy data

- total cases for three states

```r
(today = covid %>% 
  filter(date == max(date), state %in% c("California", "Hawaii", "Nevada")) %>% 
  group_by(state) %>% 
  summarize(totCases = sum(cases)) %>% 
  mutate(state = as.factor(state)))
```

```
# A tibble: 3 x 2
  state      totCases
  <fct>         <dbl>
1 California  3822073
2 Hawaii        36646
3 Nevada       334763
```

---

# Factor Manipulation

```r
# Reorder based on characteristic
fct_reorder(today$state, today$totCases) %>%  levels()
```

```
[1] "Hawaii"     "Nevada"    
[3] "California"
```

```r
fct_reorder(today$state, -today$totCases) %>% levels()
```

```
[1] "California" "Nevada"    
[3] "Hawaii"    
```

```r
# Reorder based on preference
fct_relevel(today$state,"Nevada") %>% levels()
```

```
[1] "Nevada"     "California"
[3] "Hawaii"    
```

```r
fct_recode(today$state, "NV" = "Nevada", "CA" = "California") %>% levels()
```

```
[1] "CA"     "Hawaii" "NV"    
```

---
# Recode

```r
covid_region %>% 
  mutate(region = fct_recode(region, "SOUTH" = "South")) %>% 
  ggplot(aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
  theme(legend.position = "NA") 
```

---
# Reoder By Choice

```r
covid_region %>% 
  mutate(region = fct_relevel(region, c("South"))) %>% 
  ggplot(aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
  theme(legend.position = "NA") 
```

<img src="lecture-08_files/figure-html/unnamed-chunk-19-1.png" width="432" />
---
# Reorder By Frequency of Occurance

```r
covid_region %>% 
  mutate(region = fct_infreq(region)) %>% 
  ggplot(aes(x = date, y = value)) +
  geom_line(aes(col = region)) +
  facet_grid(name~region, scale = "free_y") +
  theme_linedraw() +
  theme(legend.position = "NA")
```

---

# Dates

- Date vectors are built on double (numeric) vectors.

- They have class “Date” and no other attributes

- This is why operations like `min`, `max`, `>`, `-` work

```r
x = "2020-08-04"; y = as.Date(x)
class(x)
```

```
[1] "character"
```

```r
y = as.Date(x)
class(y)
```

```
[1] "Date"
```

```r
typeof(y)
```

```
[1] "double"
```

```r
attributes(y)
```

```
$class
[1] "Date"
```

```r
unclass(y)
```

```
[1] 18478
```
---

# Dates

The numeric value represents the `number of days` since 1970-01-01
1970-01-01 is the Unix Epoch.

```r
x = as.Date("1970-01-02")
unclass(x)
```

```
[1] 1
```

```r
x = as.Date("1969-12-31")
unclass(x)
```

```
[1] -1
```

---

# Beyond Day-Dates

- R provides two ways of storing date-time information, POSIXct, and POSIXlt. 
  - “POSIX” = Portable Operating System Interface
  - “ct”  = calendar time 
  - “lt”  = local time. Here we’ll focus on 
- `POSIXct` is most appropriate for use in data frames. 
- `POSIXct` vectors are built on top of double/numeirc, 
- The numeric value represents the number of _seconds_ since the Unix Epoch (1970-01-01).

```r
(x =  as.POSIXct("2020-08-02 22:00", tz = 'UTC'))
```

```
[1] "2020-08-02 22:00:00 UTC"
```

```r
unclass(x)
```

```
[1] 1596405600
attr(,"tzone")
[1] "UTC"
```

```r
structure(x, tzone = "America/Chicago")
```

```
[1] "2020-08-02 17:00:00 CDT"
```

```r
structure(x, tzone = "America/New_York")
```

```
[1] "2020-08-02 18:00:00 EDT"
```

```r
structure(x, tzone = "America/Los_Angeles")
```

```
[1] "2020-08-02 15:00:00 PDT"
```

---

# The basics and beyond <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg">  <path d="M505.12019,19.09375c-1.18945-5.53125-6.65819-11-12.207-12.1875C460.716,0,435.507,0,410.40747,0,307.17523,0,245.26909,55.20312,199.05238,128H94.83772c-16.34763.01562-35.55658,11.875-42.88664,26.48438L2.51562,253.29688A28.4,28.4,0,0,0,0,264a24.00867,24.00867,0,0,0,24.00582,24H127.81618l-22.47457,22.46875c-11.36521,11.36133-12.99607,32.25781,0,45.25L156.24582,406.625c11.15623,11.1875,32.15619,13.15625,45.27726,0l22.47457-22.46875V488a24.00867,24.00867,0,0,0,24.00581,24,28.55934,28.55934,0,0,0,10.707-2.51562l98.72834-49.39063c14.62888-7.29687,26.50776-26.5,26.50776-42.85937V312.79688c72.59753-46.3125,128.03493-108.40626,128.03493-211.09376C512.07526,76.5,512.07526,51.29688,505.12019,19.09375ZM384.04033,168A40,40,0,1,1,424.05,128,40.02322,40.02322,0,0,1,384.04033,168Z"></path></svg>

- So far you have learned the basics of working with tables in R
- But there are:

- Multiple ways to do things
  - Finding help
  - Implementing existing solutions

---

# Multiple ways to solve a problem...

**Question**: How many new cases each day has there been in Santa Barbara County?

**Task**: Separating cumulative values ... (today - yesterday...)

- `?diff` (base R): Returns suitably lagged and iterated differences.

- `?lag ` (dplyr): Find the "previous" (lag()) or "next" (lead()) values in a vector.

```r
(x = c(5, 15, 19,20,22))
```

```
[1]  5 15 19 20 22
```

```r
diff(x)
```

```
[1] 10  4  1  2
```

```r
lag(x)
```

```
[1] NA  5 15 19 20
```

- For `diff` the right answers but the wrong length
- For `lag` the right lengths but the "wrong" answers

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
*covid
```
]
 
.panel2-lag-auto[

```
# A tibble: 1,488,568 x 6
   date       county state fips 
   <date>     <chr>  <chr> <chr>
 1 2020-01-21 Snoho… Wash… 53061
 2 2020-01-22 Snoho… Wash… 53061
 3 2020-01-23 Snoho… Wash… 53061
 4 2020-01-24 Cook   Illi… 17031
 5 2020-01-24 Snoho… Wash… 53061
 6 2020-01-25 Orange Cali… 06059
 7 2020-01-25 Cook   Illi… 17031
 8 2020-01-25 Snoho… Wash… 53061
 9 2020-01-26 Maric… Ariz… 04013
10 2020-01-26 Los A… Cali… 06037
# … with 1,488,558 more rows,
#   and 2 more variables:
#   cases <dbl>, deaths <dbl>
```
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
* filter(county == "Santa Barbara")
```
]
 
.panel2-lag-auto[

```
# A tibble: 478 x 6
   date       county state fips 
   <date>     <chr>  <chr> <chr>
 1 2020-03-15 Santa… Cali… 06083
 2 2020-03-16 Santa… Cali… 06083
 3 2020-03-17 Santa… Cali… 06083
 4 2020-03-18 Santa… Cali… 06083
 5 2020-03-19 Santa… Cali… 06083
 6 2020-03-20 Santa… Cali… 06083
 7 2020-03-21 Santa… Cali… 06083
 8 2020-03-22 Santa… Cali… 06083
 9 2020-03-23 Santa… Cali… 06083
10 2020-03-24 Santa… Cali… 06083
# … with 468 more rows, and 2
#   more variables:
#   cases <dbl>, deaths <dbl>
```
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
* mutate(newCases = cases - lag(cases))
```
]
 
.panel2-lag-auto[

```
# A tibble: 478 x 7
   date       county state fips 
   <date>     <chr>  <chr> <chr>
 1 2020-03-15 Santa… Cali… 06083
 2 2020-03-16 Santa… Cali… 06083
 3 2020-03-17 Santa… Cali… 06083
 4 2020-03-18 Santa… Cali… 06083
 5 2020-03-19 Santa… Cali… 06083
 6 2020-03-20 Santa… Cali… 06083
 7 2020-03-21 Santa… Cali… 06083
 8 2020-03-22 Santa… Cali… 06083
 9 2020-03-23 Santa… Cali… 06083
10 2020-03-24 Santa… Cali… 06083
# … with 468 more rows, and 3
#   more variables:
#   cases <dbl>, deaths <dbl>,
#   newCases <dbl>
```
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
* mutate(newCases2 = c(NA, diff(cases)))
```
]
 
.panel2-lag-auto[

```
# A tibble: 478 x 8
   date       county state fips 
   <date>     <chr>  <chr> <chr>
 1 2020-03-15 Santa… Cali… 06083
 2 2020-03-16 Santa… Cali… 06083
 3 2020-03-17 Santa… Cali… 06083
 4 2020-03-18 Santa… Cali… 06083
 5 2020-03-19 Santa… Cali… 06083
 6 2020-03-20 Santa… Cali… 06083
 7 2020-03-21 Santa… Cali… 06083
 8 2020-03-22 Santa… Cali… 06083
 9 2020-03-23 Santa… Cali… 06083
10 2020-03-24 Santa… Cali… 06083
# … with 468 more rows, and 4
#   more variables:
#   cases <dbl>, deaths <dbl>,
#   newCases <dbl>,
#   newCases2 <dbl>
```
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
* pivot_longer(cols = c('cases', 'newCases'))
```
]
 
.panel2-lag-auto[

```
# A tibble: 956 x 8
   date       county state fips 
   <date>     <chr>  <chr> <chr>
 1 2020-03-15 Santa… Cali… 06083
 2 2020-03-15 Santa… Cali… 06083
 3 2020-03-16 Santa… Cali… 06083
 4 2020-03-16 Santa… Cali… 06083
 5 2020-03-17 Santa… Cali… 06083
 6 2020-03-17 Santa… Cali… 06083
 7 2020-03-18 Santa… Cali… 06083
 8 2020-03-18 Santa… Cali… 06083
 9 2020-03-19 Santa… Cali… 06083
10 2020-03-19 Santa… Cali… 06083
# … with 946 more rows, and 4
#   more variables:
#   deaths <dbl>,
#   newCases2 <dbl>,
#   name <chr>, value <dbl>
```
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
* geom_col(aes(color = name, fill = name))
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_08_output-1.png" width="432" />
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
  geom_col(aes(color = name, fill = name)) +
* facet_wrap(~name)
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_09_output-1.png" width="432" />
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
  geom_col(aes(color = name, fill = name)) +
  facet_wrap(~name) +
* facet_wrap(~name, scales = "free_y")
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_10_output-1.png" width="432" />
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
  geom_col(aes(color = name, fill = name)) +
  facet_wrap(~name) +
  facet_wrap(~name, scales = "free_y") +
* theme_linedraw()
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_11_output-1.png" width="432" />
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
  geom_col(aes(color = name, fill = name)) +
  facet_wrap(~name) +
  facet_wrap(~name, scales = "free_y") +
  theme_linedraw() +
* theme(legend.position = "none")
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_12_output-1.png" width="432" />
]

---
count: false
 
#Lagging Values: 2 ways
.panel1-lag-auto[

```r
covid %>%
  filter(county == "Santa Barbara") %>%
  mutate(newCases = cases - lag(cases)) %>%
  mutate(newCases2 = c(NA, diff(cases))) %>%
  pivot_longer(cols = c('cases', 'newCases')) %>%
  mutate(name = fct_recode(name,
                           "Cummulative Cases" = "cases",
                           "New Cases" = "newCases")) %>%
  ggplot(aes(x = date, y = value)) +
  geom_col(aes(color = name, fill = name)) +
  facet_wrap(~name) +
  facet_wrap(~name, scales = "free_y") +
  theme_linedraw() +
  theme(legend.position = "none") +
* labs(title = "Santa Barbara Cases Cases")
```
]
 
.panel2-lag-auto[
<img src="lecture-08_files/figure-html/lag_auto_13_output-1.png" width="432" />
]

---
# Real World Application

- From a New York Times Article

---
# Getting Help

---
# Reading the docs ...

```r
install.packages("zoo")
library(zoo)
?rollmean
```

---
# Testing ...:

```r
(x = c(1:20))
```

```
 [1]  1  2  3  4  5  6  7  8  9
[10] 10 11 12 13 14 15 16 17 18
[19] 19 20
```

```r
#k	integer width of the rolling window. Must be odd for rollmedian.
rollmean(x, k =5)
```

```
 [1]  3  4  5  6  7  8  9 10 11
[10] 12 13 14 15 16 17 18
```

```r
#fill: a three-component vector or list (recycled otherwise) providing filling values 
#      at the left/within/to the right of the data range. 
rollmean(x, k= 5, fill = NA)
```

```
 [1] NA NA  3  4  5  6  7  8  9
[10] 10 11 12 13 14 15 16 17 18
[19] NA NA
```

```r
#align: character specifying whether the index of the result should be 
#       left- or right-aligned or centered (default) compared to the rolling 
#       window of observations.
rollmean(x, k= 5, fill = NA, align = "right")
```

```
 [1] NA NA NA NA  3  4  5  6  7
[10]  8  9 10 11 12 13 14 15 16
[19] 17 18
```

---

```r
*state.of.interest = "Alabama"
```
]
 
.panel2-al-auto[

]

---
count: false
 
#Replicate the NY-Times Image
.panel1-al-auto[