The Tidyverse

---

## The Tidyverse

### Applications of Data Science - Class 1

### Giora Simchoni

#### `gsimchoni@gmail.com and add #dsapps in subject`

### Stat. and OR Department, TAU
### 2023-02-23

---

<div class="my-footer">
  <span>
    <a href="https://dsapps-2023.github.io/Class_Slides/" target="_blank">Applications of Data Science
    </a>
  </span>
</div>

---

# I don't need to know about wrangling data, I get by.

---

# So, what's wrong with Excel?

(MS Excel is one amazing software. But it lacks:)

- Structure (or rather, structure is up to the user)
- Types to variables
- Automation (you could learn VBA Excel, but the horror)
- Reproducibility
- Open Source
- Extensibility
- Speed and Scale
- Modeling (there *is* a t-test, but the horror)

[Excel might be the most dangerous software on the planet](https://www.forbes.com/sites/timworstall/2013/02/13/microsofts-excel-might-be-the-most-dangerous-software-on-the-planet/#667084f5633d) (Forbes)

[Excel horror stories](https://eusprig.org/research-info/horror-stories/)

---

# So, what's wrong with base R?

(Base R is one amazing software. But it lacks:)

- Consistency:
  - Function names
  - Function arguments names
  - Function arguments order
  - Function return types (sometimes the same function!)
- Meaningful errors and warnings
- Good choices of default values to arguments
- Speed
- Good and easy visualizations
- One other thing

---

### (In) Consistency - Example 1: Strings

```r
# split a string by pattern: strsplit(string, pattern)
strsplit("Who dis?", " ")
```

```
## [[1]]
## [1] "Who"  "dis?"
```

```r
# find if a pattern exists in a string: grepl(pattern, string)
grepl("di", "Who dis?")
```

```
## [1] TRUE
```

```r
# substitute a pattern in a string: sub(pattern, replace, string)
sub("di", "thi", "Who dis?")
```

```
## [1] "Who this?"
```

```r
# length of a string: nchar(string); length of object: length(obj)
c(nchar("Who dis?"), length("Who dis?"))
```

```
## [1] 8 1
```
]

---

### (In) Consistency - Example 2: Models

```r
n <- 10000
x1 <- runif(n)
x2 <- runif(n)
t <- 1 + 2 * x1 + 3 * x2
y <- rbinom(n, 1, 1 / (1 + exp(-t)))
```

```r
glm(y ~ x1 + x2, family = "binomial")
```

```r
glmnet(as.matrix(cbind(x1, x2)), as.factor(y), family = "binomial")
```

```r
randomForest(as.factor(y) ~ x1 + x2)
```

```r
gbm(y ~ x1 + x2, data = data.frame(x1 = x1, x2 = x2, y = y))
```

😱

---

### (Un) Meaningful Errors - Example

```r
df <- data.frame(Education = 1:5, Ethnicity = c(2, 4, 5, 2, 1))
table(df$Eduction, df$Ethnicity)
```

<pre style="color: red;"><code>## Error in table(df$Eduction, df$Ethnicity): all arguments must have the same
length
</code></pre>

---

### (Bad) Default Values - Example

![](images/bad_args_table.png)

```r
# In R 3.6... in R 4.0 this was fixed!
df <- read.csv("../data/bad_args_test.csv")
df$col3
```

```
## [1] a b c d
## Levels: a b c d
```

```r
df <- read.csv("../data/bad_args_test.csv", stringsAsFactors = FALSE)
df$col3
```

```
## [1] "a" "b" "c" "d"
```

---

### (No) Speed - Example

```r
file_path <- "../data/mediocre_file.csv"
df <- read.csv(file_path)
dim(df)
```

```
## [1] 9180   14
```

```r
library(microbenchmark)

microbenchmark(
  read_base = read.csv(file_path),
  read_tidy = read_csv(file_path, col_types = cols()),
  read_dt = data.table::fread(file_path),
  times = 10)
```

```
## Unit: milliseconds
##       expr       min        lq     mean   median        uq      max neval
##  read_base 42.623101 43.625400 45.69589 44.70140 45.254901  56.8816    10
##  read_tidy 28.173101 28.844001 40.23915 30.37425 31.458502 131.9949    10
##    read_dt  8.418402  8.626601 15.49107  8.81210  9.441102  74.9264    10
```

---
class: section-slide

# Detour: The OKCupid Dataset

---

## The OKCupid Dataset

- ~60K active OKCupid users scraped on June 2012
- 35K Male, 25K Female (less awareness for non-binary back then)
- Answers to questions like:
  - Body Type
  - Diet
  - Substance Abuse
  - Education
  - Do you like pets?
  - Open questions, e.g. "On a typical Friday night I am..."
  - And the more boring demographic details like age, height, location, sign, religion etc.
- See [here](https://github.com/rudeboybert/JSE_OkCupid/blob/master/okcupid_codebook_revised.txt) for the full codebook

---

### BTW

See [here](https://www.tandfonline.com/doi/full/10.1080/26939169.2021.1930812).

---
class: section-slide

# End of Detour

---

### (Not) Good Vizualizations - Example

```r
okcupid <- read_csv("~/okcupid.csv.zip", col_types = cols())
okcupid$income[okcupid$income == -1] <- NA
okcupid$height_cm <- okcupid$height * 2.54
```

```r
plot(okcupid$height_cm, log10(okcupid$income + 1),
     col = c("red", "green")[as.factor(okcupid$sex)])
```

<img src="images/Viz-Base-1.png" width="70%" />
]

---

```r
ggplot(okcupid, aes(height_cm, log10(income + 1), color = sex)) +
  geom_point()
```

<pre style="color: red;"><code>## Warning: Removed 48442 rows containing missing values (`geom_point()`).
</code></pre><img src="images/Viz-Tidy-1.png" width="70%" />

---

## One other thing

Manager: "Give me the average income of women respondents above age 30 grouped by sexual orientation!"

You:

```r
mean_bi <- mean(okcupid$income[okcupid$sex == "f" & okcupid$age > 30 & okcupid$orientation == "bisexual"], na.rm = TRUE)
mean_gay <- mean(okcupid$income[okcupid$sex == "f" & okcupid$age > 30 & okcupid$orientation == "gay"], na.rm = TRUE)
mean_straight <- mean(okcupid$income[okcupid$sex == "f" & okcupid$age > 30 & okcupid$orientation == "straight"], na.rm = TRUE)

data.frame(orientation = c("bisexual", "gay", "straight"),
           income_mean = c(mean_bi, mean_gay, mean_straight))
```

```
##   orientation income_mean
## 1    bisexual   133421.05
## 2         gay    86489.36
## 3    straight    85219.74
```

---

Or the slightly better you:

```r
mean_income_function <- function(orientation) {
  mean(okcupid$income[okcupid$sex == "f" & okcupid$age > 30 & okcupid$orientation == orientation], na.rm = TRUE)
}

mean_bi <- mean_income_function("bisexual")
mean_gay <- mean_income_function("gay")
mean_straight <- mean_income_function("straight")

data.frame(orientation = c("bisexual", "gay", "straight"),
           income_mean = c(mean_bi, mean_gay, mean_straight))
```

```
##   orientation income_mean
## 1    bisexual   133421.05
## 2         gay    86489.36
## 3    straight    85219.74
```

---

Or the even better you:

```r
orientations <- c("bisexual", "gay", "straight")
income_means <- numeric(3)

for (i in seq_along(orientations)) {
  income_means[i] <- mean_income_function(orientations[i])
}

data.frame(orientation = orientations, income_mean = income_means)
```

```
##   orientation income_mean
## 1    bisexual   133421.05
## 2         gay    86489.36
## 3    straight    85219.74
```

---

Or the best you:

```r
okcupid_females_over30 <- with(okcupid, okcupid[sex == "f" & age > 30, ])
aggregate(okcupid_females_over30$income,
          by = list(orientation = okcupid_females_over30$orientation),
          FUN = mean, na.rm = TRUE)
```

```
##   orientation         x
## 1    bisexual 133421.05
## 2         gay  86489.36
## 3    straight  85219.74
```

<br>

Manager: "What? Why would bisexual women have a higher income than straight or gay women? Could you add the median, trimmed mean, standard error and n?"

You: 😱

---
class: section-slide

# The Tidyverse

---

## What *is* The Tidyverse?

> The [tidyverse](https://www.tidyverse.org/) is an opinionated collection of R packages designed for data science. All packages share an underlying design philosophy, grammar, and data structures.

- `tibble`: the `data.frame` re-imagined
- `readr`: importing/exporting (mostly rectangular) data for humans
- `dplyr` + `tidyr`: a grammar of data manipulation
- `purrr`: functional programming in R
- `stringr`: string manipulation
- `ggplot2`: a grammar of graphics

---

The above can all be installed and loaded under the `tidyverse` package:

```r
library(tidyverse)
```

Many more:
- `lubridate`: manipulating dates
- `tidymodels`: tidy modeling/statistics
- `rvest`: web scraping
- `tidytext`: tidy text analysis (life saver)
- `tidygraph` + `ggraph`: manipulating and plotting networks
- `glue`: print like a boss
- countless `gg` extensions (`ggmosaic`, `ggbeeswarm`, `gganimate`, `ggridges` etc.)

---

## What's so great about the Tidyverse?

- Tidy Data
- Consistency (in function names, args, return types, documentation)
- The Pipe
- Speed (C++ under the hood)
- `ggplot2`
- The Community

---

## Tidy Data

- Each variable must have its own column.
- Each observation must have its own row.
- Each value must have its own cell.

<br>

---

### Which one of these datasets is tidy? (I)

```r
table1
```

```
## # A tibble: 315 × 4
##    religion      yob n_straight n_total
##    <chr>       <dbl>      <dbl>   <dbl>
##  1 atheist      1950         26      29
##  2 buddhist     1950          6       6
##  3 christian    1950         28      32
##  4 hindu        1950          0       0
##  5 jewish       1950         21      24
##  6 muslim       1950          0       0
##  7 unspecified  1950         71      76
##  8 atheist      1951         31      33
##  9 buddhist     1951         11      11
## 10 christian    1951         23      24
## # … with 305 more rows
```

---

### Which one of these datasets is tidy? (II)

```r
table2
```

```
## # A tibble: 630 × 4
##    religion    yob type         n
##    <chr>     <dbl> <chr>    <dbl>
##  1 atheist    1950 straight    26
##  2 atheist    1950 total       29
##  3 buddhist   1950 straight     6
##  4 buddhist   1950 total        6
##  5 christian  1950 straight    28
##  6 christian  1950 total       32
##  7 hindu      1950 straight     0
##  8 hindu      1950 total        0
##  9 jewish     1950 straight    21
## 10 jewish     1950 total       24
## # … with 620 more rows
```

---

### Which one of these datasets is tidy? (III)

```r
table3
```

```
## # A tibble: 315 × 3
##    religion      yob pct_straight
##    <chr>       <dbl> <chr>       
##  1 atheist      1950 26/29       
##  2 buddhist     1950 6/6         
##  3 christian    1950 28/32       
##  4 hindu        1950 0/0         
##  5 jewish       1950 21/24       
##  6 muslim       1950 0/0         
##  7 unspecified  1950 71/76       
##  8 atheist      1951 31/33       
##  9 buddhist     1951 11/11       
## 10 christian    1951 23/24       
## # … with 305 more rows
```

---

### Which one of these datasets is tidy? (IV)

```r
table4
```

```
## # A tibble: 7 × 91
##   religion    n_total_1950 n_total_1951 n_total_1952 n_total_1953 n_total_1954
##   <chr>              <dbl>        <dbl>        <dbl>        <dbl>        <dbl>
## 1 atheist               29           33           34           37           40
## 2 buddhist               6           11           14           16           11
## 3 christian             32           24           37           47           37
## 4 hindu                  0            0            0            1            1
## 5 jewish                24           29           27           23           25
## 6 muslim                 0            0            0            0            0
## 7 unspecified           76           79           83           97           83
## # … with 85 more variables: n_total_1955 <dbl>, n_total_1956 <dbl>,
## #   n_total_1957 <dbl>, n_total_1958 <dbl>, n_total_1959 <dbl>,
## #   n_total_1960 <dbl>, n_total_1961 <dbl>, n_total_1962 <dbl>,
## #   n_total_1963 <dbl>, n_total_1964 <dbl>, n_total_1965 <dbl>,
## #   n_total_1966 <dbl>, n_total_1967 <dbl>, n_total_1968 <dbl>,
## #   n_total_1969 <dbl>, n_total_1970 <dbl>, n_total_1971 <dbl>,
## #   n_total_1972 <dbl>, n_total_1973 <dbl>, n_total_1974 <dbl>, …
```

---

### Why Tidy?

> Happy families are all alike; every unhappy family is unhappy in its own way. (Leo Tolstoy)

<br>

> It allows R’s vectorised nature to shine. (Hadley Wickham)

---

### A Tidy dataset will be much easier to transform

```r
table1$pct_straight = table1$n_straight / table1$n_total
table1
```

```
## # A tibble: 315 × 5
##    religion      yob n_straight n_total pct_straight
##    <chr>       <dbl>      <dbl>   <dbl>        <dbl>
##  1 atheist      1950         26      29        0.897
##  2 buddhist     1950          6       6        1    
##  3 christian    1950         28      32        0.875
##  4 hindu        1950          0       0      NaN    
##  5 jewish       1950         21      24        0.875
##  6 muslim       1950          0       0      NaN    
##  7 unspecified  1950         71      76        0.934
##  8 atheist      1951         31      33        0.939
##  9 buddhist     1951         11      11        1    
## 10 christian    1951         23      24        0.958
## # … with 305 more rows
```

---

### A Tidy dataset will be much easier to plot

```r
ggplot(table1, aes(x = yob, y = pct_straight, color = religion)) +
  geom_smooth(method = "loess", formula = y ~ x, se = FALSE)
```

---
class: section-slide

# Detour: The `tibble`

---

### The `tibble`: the `data.frame` re-imagined

- Prints nicer:

```r
tib1 <- tibble(day = lubridate::today() + runif(1e3) * 30,
               type = sample(letters, 1e3, replace = TRUE),
               quantity = sample(seq(0, 100, 10), 1e3, replace = TRUE))
tib1
```

```
## # A tibble: 1,000 × 3
##    day        type  quantity
##    <date>     <chr>    <dbl>
##  1 2023-03-17 j           90
##  2 2023-03-15 o           40
##  3 2023-02-28 q           60
##  4 2023-02-28 c            0
##  5 2023-03-11 v           90
##  6 2023-03-06 s           30
##  7 2023-03-23 h           50
##  8 2023-02-25 q           10
##  9 2023-03-04 x           70
## 10 2023-03-23 m           90
## # … with 990 more rows
```

---

```r
df1 <- data.frame(day = lubridate::today() + runif(1e3) * 30,
                  type = sample(letters, 1e3, replace = TRUE),
                  quantity = sample(seq(0, 100, 10), 1e3, replace = TRUE))
df1
```

```
##             day type quantity
## 1    2023-03-07    d       80
## 2    2023-03-07    u       90
## 3    2023-03-02    c       60
## 4    2023-03-11    h        0
## 5    2023-03-01    j       90
## 6    2023-03-21    t      100
## 7    2023-03-06    g       70
## 8    2023-03-09    r        0
## 9    2023-03-24    n       70
## 10   2023-03-22    f       40
## 11   2023-03-18    a       70
## 12   2023-03-19    n      100
## 13   2023-03-15    o       10
## 14   2023-03-20    u       60
## 15   2023-03-07    i       70
## 16   2023-03-12    o       50
## 17   2023-03-07    h       70
## 18   2023-03-11    r       70
## 19   2023-03-18    r       50
## 20   2023-03-07    h       20
## 21   2023-03-05    i       30
## 22   2023-03-20    k       70
## 23   2023-03-10    s       80
## 24   2023-03-16    y       70
## 25   2023-02-25    e       40
## 26   2023-03-23    c       30
## 27   2023-03-10    m      100
## 28   2023-03-08    a      100
## 29   2023-03-21    z       90
## 30   2023-02-28    d       50
## 31   2023-03-16    u       40
## 32   2023-03-02    o      100
## 33   2023-03-04    l       60
## 34   2023-03-08    n       80
## 35   2023-02-26    s       40
## 36   2023-03-12    v       60
## 37   2023-03-13    u      100
## 38   2023-03-04    h       10
## 39   2023-03-07    x       20
## 40   2023-03-23    l       60
## 41   2023-03-07    h       70
## 42   2023-03-24    h       40
## 43   2023-03-11    h       70
## 44   2023-03-15    i        0
## 45   2023-03-05    h       50
## 46   2023-02-25    c       60
## 47   2023-02-24    l       20
## 48   2023-03-12    k       50
## 49   2023-03-10    t       70
## 50   2023-03-08    u       30
## 51   2023-03-04    y       40
## 52   2023-03-21    h       30
## 53   2023-03-18    p       50
## 54   2023-03-20    d       70
## 55   2023-03-24    f        0
## 56   2023-02-24    k       20
## 57   2023-03-16    q       30
## 58   2023-02-25    x       70
## 59   2023-03-23    p       10
## 60   2023-03-17    g       70
## 61   2023-03-23    a       70
## 62   2023-03-05    n       30
## 63   2023-03-18    p       30
## 64   2023-03-15    n       60
## 65   2023-03-08    n        0
## 66   2023-02-27    j       10
## 67   2023-03-08    k      100
## 68   2023-03-18    h      100
## 69   2023-03-02    b       80
## 70   2023-03-22    t       70
## 71   2023-03-08    a       10
## 72   2023-03-07    q       50
## 73   2023-03-15    l        0
## 74   2023-02-23    g       60
## 75   2023-03-12    e       30
## 76   2023-03-19    e       20
## 77   2023-03-19    b       90
## 78   2023-03-14    e       90
## 79   2023-03-15    c       20
## 80   2023-03-23    c       10
## 81   2023-03-19    o       30
## 82   2023-03-12    z        0
## 83   2023-03-18    t      100
## 84   2023-03-09    s       60
## 85   2023-03-11    o       10
## 86   2023-02-26    h       30
## 87   2023-03-23    f       20
## 88   2023-03-07    r        0
## 89   2023-03-19    o      100
## 90   2023-03-01    p       20
## 91   2023-03-01    i      100
## 92   2023-03-18    a       10
## 93   2023-03-06    b       90
## 94   2023-03-20    o       60
## 95   2023-03-15    v       70
## 96   2023-03-22    j       60
## 97   2023-03-01    k       10
## 98   2023-03-10    r       90
## 99   2023-03-22    q       70
## 100  2023-03-01    p       20
## 101  2023-03-03    w       80
## 102  2023-03-17    o       80
## 103  2023-03-21    f       40
## 104  2023-03-05    s       60
## 105  2023-02-28    z       30
## 106  2023-03-12    u       80
## 107  2023-03-05    j       90
## 108  2023-03-11    v       10
## 109  2023-03-10    k        0
## 110  2023-02-24    o       70
## 111  2023-02-25    s       40
## 112  2023-03-13    x       20
## 113  2023-03-04    z       60
## 114  2023-02-24    m      100
## 115  2023-03-17    q       10
## 116  2023-03-15    g       50
## 117  2023-03-24    x      100
## 118  2023-02-27    v       10
## 119  2023-02-27    y       70
## 120  2023-03-06    d       40
## 121  2023-02-24    a       30
## 122  2023-03-16    p       70
## 123  2023-03-10    c      100
## 124  2023-02-28    x       90
## 125  2023-02-26    s       70
## 126  2023-03-21    v       90
## 127  2023-03-16    x       40
## 128  2023-02-23    h       40
## 129  2023-03-21    y       70
## 130  2023-02-24    u      100
## 131  2023-02-26    v       40
## 132  2023-03-22    h       10
## 133  2023-03-15    k       60
## 134  2023-02-27    k       30
## 135  2023-02-24    m       30
## 136  2023-03-04    j      100
## 137  2023-02-25    p       40
## 138  2023-03-08    r       90
## 139  2023-02-28    a        0
## 140  2023-03-23    l        0
## 141  2023-03-12    n      100
## 142  2023-03-02    c       60
## 143  2023-03-13    s       80
## 144  2023-03-17    h      100
## 145  2023-03-11    v       80
## 146  2023-03-18    x       80
## 147  2023-03-20    q      100
## 148  2023-03-05    j       30
## 149  2023-03-13    n       70
## 150  2023-03-14    z      100
## 151  2023-03-20    f        0
## 152  2023-03-03    v       60
## 153  2023-02-24    j      100
## 154  2023-02-24    q       50
## 155  2023-03-18    h       60
## 156  2023-03-02    d       30
## 157  2023-03-15    v       40
## 158  2023-03-13    f      100
## 159  2023-03-21    w       40
## 160  2023-03-17    r       30
## 161  2023-02-23    h       30
## 162  2023-03-21    s        0
## 163  2023-03-02    v       10
## 164  2023-03-04    h       80
## 165  2023-02-28    e      100
## 166  2023-03-12    z       90
## 167  2023-03-16    q       10
## 168  2023-03-17    j       20
## 169  2023-03-14    i       70
## 170  2023-03-12    n       50
## 171  2023-03-15    h       10
## 172  2023-03-03    b       30
## 173  2023-03-15    f       50
## 174  2023-03-13    f       80
## 175  2023-03-03    g        0
## 176  2023-03-18    d       20
## 177  2023-03-20    e      100
## 178  2023-03-13    x       80
## 179  2023-03-10    s       40
## 180  2023-03-23    a       20
## 181  2023-03-11    q       50
## 182  2023-03-01    s       40
## 183  2023-03-17    y       80
## 184  2023-03-23    y       10
## 185  2023-03-16    w       10
## 186  2023-02-24    v       10
## 187  2023-03-01    k       20
## 188  2023-03-24    b       80
## 189  2023-03-03    p       30
## 190  2023-03-18    d       10
## 191  2023-03-24    i       30
## 192  2023-02-26    h       60
## 193  2023-03-20    y       20
## 194  2023-03-13    d       40
## 195  2023-03-12    s       90
## 196  2023-02-23    v       80
## 197  2023-02-28    w      100
## 198  2023-03-02    v       90
## 199  2023-03-19    i        0
## 200  2023-03-11    s       10
## 201  2023-03-09    v       40
## 202  2023-02-25    h        0
## 203  2023-03-11    h       40
## 204  2023-03-12    s      100
## 205  2023-02-25    o       40
## 206  2023-03-18    e       10
## 207  2023-03-15    b        0
## 208  2023-03-21    n       60
## 209  2023-02-27    s       10
## 210  2023-02-25    a       40
## 211  2023-02-27    e       40
## 212  2023-02-25    x       20
## 213  2023-03-24    u      100
## 214  2023-03-05    j       90
## 215  2023-03-18    z        0
## 216  2023-03-06    h       10
## 217  2023-03-09    y       50
## 218  2023-03-02    p       50
## 219  2023-02-26    t       10
## 220  2023-03-19    o       50
## 221  2023-03-15    q       30
## 222  2023-02-25    l      100
## 223  2023-02-23    v       40
## 224  2023-02-28    m       60
## 225  2023-03-04    o       10
## 226  2023-03-18    u        0
## 227  2023-03-03    h       60
## 228  2023-02-23    f       90
## 229  2023-02-27    a       30
## 230  2023-03-04    a       60
## 231  2023-03-24    e       10
## 232  2023-03-20    c       20
## 233  2023-03-03    i       70
## 234  2023-03-07    p       40
## 235  2023-02-27    m       40
## 236  2023-03-04    f       70
## 237  2023-03-18    k       10
## 238  2023-03-09    l      100
## 239  2023-03-18    n       70
## 240  2023-03-16    u       50
## 241  2023-03-12    w       80
## 242  2023-03-19    k       30
## 243  2023-03-20    g        0
## 244  2023-03-03    x      100
## 245  2023-03-10    n       20
## 246  2023-03-11    t       30
## 247  2023-03-23    g       30
## 248  2023-03-10    d       20
## 249  2023-02-24    f       20
## 250  2023-03-20    f       40
## 251  2023-03-04    u       80
## 252  2023-03-18    u       70
## 253  2023-03-06    s       50
## 254  2023-03-14    r       50
## 255  2023-03-24    i       50
## 256  2023-03-18    n       60
## 257  2023-03-08    q        0
## 258  2023-03-03    t       40
## 259  2023-03-12    b       10
## 260  2023-03-01    p       50
## 261  2023-03-22    w       90
## 262  2023-02-24    r       60
## 263  2023-03-13    w       10
## 264  2023-03-22    s       60
## 265  2023-03-08    a       40
## 266  2023-03-16    f      100
## 267  2023-03-13    d       80
## 268  2023-02-25    k       60
## 269  2023-02-25    b       10
## 270  2023-03-21    o       90
## 271  2023-03-06    c       80
## 272  2023-02-23    w        0
## 273  2023-03-07    q       70
## 274  2023-03-13    y       80
## 275  2023-03-11    w       90
## 276  2023-03-08    q        0
## 277  2023-03-16    e        0
## 278  2023-02-25    g       20
## 279  2023-03-12    u        0
## 280  2023-03-22    e       90
## 281  2023-03-12    w       30
## 282  2023-03-07    h       40
## 283  2023-03-20    r       80
## 284  2023-03-15    d      100
## 285  2023-03-22    p       60
## 286  2023-03-03    w       30
## 287  2023-03-02    d       60
## 288  2023-03-05    f       50
## 289  2023-02-24    x       50
## 290  2023-02-24    i       90
## 291  2023-03-13    u       50
## 292  2023-03-09    g       10
## 293  2023-02-27    g       50
## 294  2023-03-14    q       70
## 295  2023-03-03    w       20
## 296  2023-03-16    z       30
## 297  2023-03-20    j       80
## 298  2023-03-24    b       90
## 299  2023-03-21    k       50
## 300  2023-03-20    z       50
## 301  2023-03-12    q       50
## 302  2023-03-16    v        0
## 303  2023-03-21    d       40
## 304  2023-03-23    d       50
## 305  2023-02-28    o       40
## 306  2023-03-05    e       50
## 307  2023-03-02    t       30
## 308  2023-03-02    m        0
## 309  2023-03-06    r       60
## 310  2023-03-17    e       90
## 311  2023-03-23    h       20
## 312  2023-03-06    p      100
## 313  2023-03-23    k       60
## 314  2023-03-01    t      100
## 315  2023-03-12    g      100
## 316  2023-03-20    k        0
## 317  2023-03-20    r       30
## 318  2023-03-07    p       50
## 319  2023-03-21    i       90
## 320  2023-02-26    t        0
## 321  2023-03-08    w      100
## 322  2023-03-09    y       30
## 323  2023-03-08    q       30
## 324  2023-02-26    m       60
## 325  2023-03-04    t        0
## 326  2023-03-24    d       30
## 327  2023-03-08    n       10
## 328  2023-03-03    a       50
## 329  2023-02-23    v       80
## 330  2023-02-26    e       50
## 331  2023-03-06    q        0
## 332  2023-03-03    d       40
## 333  2023-02-28    i       10
## 334  2023-02-23    g       30
## 335  2023-03-08    o       30
## 336  2023-03-03    u       90
## 337  2023-03-02    o       10
## 338  2023-03-13    t       40
## 339  2023-03-11    a       10
## 340  2023-03-12    r        0
## 341  2023-03-18    p       10
## 342  2023-03-03    j      100
## 343  2023-02-25    w       90
## 344  2023-02-23    s       30
## 345  2023-02-26    f       70
## 346  2023-03-05    r       30
## 347  2023-03-23    v       60
## 348  2023-03-19    y       10
## 349  2023-02-25    s       90
## 350  2023-02-27    z       40
## 351  2023-03-19    x       30
## 352  2023-03-06    k       40
## 353  2023-02-24    h        0
## 354  2023-03-16    h       90
## 355  2023-03-24    h       90
## 356  2023-03-14    a       80
## 357  2023-03-12    p       40
## 358  2023-03-01    a       90
## 359  2023-03-20    j        0
## 360  2023-03-19    e       60
## 361  2023-03-04    x       10
## 362  2023-03-20    c       40
## 363  2023-03-24    o       10
## 364  2023-03-11    h       20
## 365  2023-03-06    f       60
## 366  2023-03-24    c       10
## 367  2023-03-02    s       60
## 368  2023-03-18    x        0
## 369  2023-02-25    o       20
## 370  2023-03-15    t       60
## 371  2023-02-24    o       40
## 372  2023-03-06    s       30
## 373  2023-03-20    l       80
## 374  2023-03-07    m        0
## 375  2023-03-02    r       60
## 376  2023-03-17    b       90
## 377  2023-03-16    a      100
## 378  2023-03-06    j       80
## 379  2023-03-21    o       40
## 380  2023-03-17    d       50
## 381  2023-02-25    f       30
## 382  2023-02-27    k       40
## 383  2023-03-17    o       40
## 384  2023-03-21    m       20
## 385  2023-02-25    h       50
## 386  2023-02-28    c        0
## 387  2023-03-13    e       80
## 388  2023-03-14    o       20
## 389  2023-03-08    q       70
## 390  2023-03-11    m       50
## 391  2023-03-02    i       60
## 392  2023-03-11    t       20
## 393  2023-03-14    m        0
## 394  2023-03-08    n       90
## 395  2023-02-27    c       80
## 396  2023-03-12    o       30
## 397  2023-03-17    q       90
## 398  2023-03-02    t      100
## 399  2023-03-07    b       60
## 400  2023-03-17    y       50
## 401  2023-03-01    j       60
## 402  2023-02-23    h        0
## 403  2023-03-06    a       50
## 404  2023-03-05    o       60
## 405  2023-03-17    n       40
## 406  2023-02-25    u       90
## 407  2023-03-16    d      100
## 408  2023-03-22    z       80
## 409  2023-02-28    q       40
## 410  2023-03-11    d       40
## 411  2023-03-19    q        0
## 412  2023-03-02    r       80
## 413  2023-03-21    a       90
## 414  2023-02-27    m       20
## 415  2023-03-19    v      100
## 416  2023-03-04    m       10
## 417  2023-03-10    i      100
## 418  2023-03-08    n       60
## 419  2023-03-05    q       20
## 420  2023-03-17    y       20
## 421  2023-03-18    y       70
## 422  2023-03-02    u       60
## 423  2023-03-12    k       40
## 424  2023-03-16    n      100
## 425  2023-02-27    r       30
## 426  2023-03-07    l       40
## 427  2023-03-05    i       80
## 428  2023-02-27    i       30
## 429  2023-03-03    q       20
## 430  2023-03-18    n       30
## 431  2023-03-17    y       30
## 432  2023-02-23    f        0
## 433  2023-03-23    l      100
## 434  2023-03-04    t       80
## 435  2023-02-27    l       50
## 436  2023-02-27    g       70
## 437  2023-02-26    n       20
## 438  2023-03-18    p       20
## 439  2023-02-27    p       50
## 440  2023-03-23    w       50
## 441  2023-03-02    t       20
## 442  2023-03-21    u       30
## 443  2023-03-19    o       40
## 444  2023-03-08    i       90
## 445  2023-03-15    s        0
## 446  2023-03-12    t       40
## 447  2023-03-01    j      100
## 448  2023-02-25    j        0
## 449  2023-03-21    c       20
## 450  2023-03-13    h       80
## 451  2023-02-23    x       60
## 452  2023-03-08    x      100
## 453  2023-03-12    a       80
## 454  2023-03-06    d       60
## 455  2023-03-21    l       90
## 456  2023-03-02    q       70
## 457  2023-03-05    c       70
## 458  2023-03-09    h       70
## 459  2023-03-19    z       90
## 460  2023-03-21    d       80
## 461  2023-03-12    t       80
## 462  2023-03-08    d      100
## 463  2023-02-26    s       60
## 464  2023-03-12    h        0
## 465  2023-03-03    d       40
## 466  2023-03-03    g       90
## 467  2023-03-18    g       80
## 468  2023-02-27    g       40
## 469  2023-03-14    c       40
## 470  2023-03-20    f       40
## 471  2023-02-25    n      100
## 472  2023-03-10    z        0
## 473  2023-02-23    s       80
## 474  2023-03-17    l       40
## 475  2023-03-01    m       70
## 476  2023-03-04    q        0
## 477  2023-03-22    n        0
## 478  2023-03-22    b        0
## 479  2023-03-16    o        0
## 480  2023-02-27    j       30
## 481  2023-03-05    s      100
## 482  2023-03-10    e       30
## 483  2023-03-13    z       60
## 484  2023-03-04    f        0
## 485  2023-03-01    g        0
## 486  2023-02-25    i       60
## 487  2023-03-11    e       60
## 488  2023-03-12    v       60
## 489  2023-03-20    d       90
## 490  2023-02-27    g       70
## 491  2023-02-24    i       80
## 492  2023-03-23    h       30
## 493  2023-02-25    v        0
## 494  2023-03-11    i       10
## 495  2023-02-28    n       90
## 496  2023-03-05    m       90
## 497  2023-03-12    e       60
## 498  2023-03-17    j       10
## 499  2023-03-13    s       50
## 500  2023-03-08    a       90
## 501  2023-03-23    r       60
## 502  2023-03-06    m       50
## 503  2023-03-24    l       60
## 504  2023-02-23    r       20
## 505  2023-02-24    d       30
## 506  2023-03-22    t        0
## 507  2023-03-11    a       90
## 508  2023-03-16    b       30
## 509  2023-03-11    a       10
## 510  2023-03-20    v       60
## 511  2023-03-07    j       10
## 512  2023-02-28    q       70
## 513  2023-03-23    t       10
## 514  2023-03-23    j       40
## 515  2023-03-04    z        0
## 516  2023-02-28    d       80
## 517  2023-03-24    f       50
## 518  2023-03-11    q       10
## 519  2023-03-24    j       60
## 520  2023-02-25    a       80
## 521  2023-03-13    v       70
## 522  2023-03-08    d       30
## 523  2023-03-04    g       10
## 524  2023-02-24    x       90
## 525  2023-03-13    a       50
## 526  2023-03-12    i       10
## 527  2023-03-05    j       60
## 528  2023-03-06    g      100
## 529  2023-02-23    r       70
## 530  2023-03-05    k       30
## 531  2023-03-08    e       40
## 532  2023-02-23    z       80
## 533  2023-03-18    h       80
## 534  2023-03-02    v       80
## 535  2023-03-16    v        0
## 536  2023-02-27    z       50
## 537  2023-03-23    n       10
## 538  2023-02-24    e       90
## 539  2023-02-27    h       20
## 540  2023-03-04    k       90
## 541  2023-03-11    r       90
## 542  2023-02-26    m       30
## 543  2023-03-16    g       90
## 544  2023-03-21    r       50
## 545  2023-03-11    a       10
## 546  2023-02-27    k        0
## 547  2023-02-25    f       20
## 548  2023-03-18    a       10
## 549  2023-03-02    y       80
## 550  2023-02-25    k       70
## 551  2023-03-09    n       20
## 552  2023-03-03    f       50
## 553  2023-02-25    q       90
## 554  2023-03-11    g      100
## 555  2023-03-02    j       60
## 556  2023-03-04    j      100
## 557  2023-03-15    c       60
## 558  2023-02-24    z       60
## 559  2023-03-05    v      100
## 560  2023-03-09    p       20
## 561  2023-02-25    g       40
## 562  2023-03-01    o       80
## 563  2023-03-08    v       20
## 564  2023-03-05    n       10
## 565  2023-02-27    l       10
## 566  2023-02-28    p       50
## 567  2023-03-15    p       20
## 568  2023-03-14    v      100
## 569  2023-02-23    z        0
## 570  2023-03-09    n       40
## 571  2023-03-09    t       20
## 572  2023-03-08    g       70
## 573  2023-03-02    r       90
## 574  2023-03-24    p       90
## 575  2023-03-14    c       50
## 576  2023-03-11    i       90
## 577  2023-03-11    f       20
## 578  2023-02-23    l       90
## 579  2023-03-02    b       80
## 580  2023-03-15    t       30
## 581  2023-02-24    l       90
## 582  2023-03-10    a       10
## 583  2023-03-08    i       60
## 584  2023-03-09    x       40
## 585  2023-03-14    y       40
## 586  2023-03-01    a       90
## 587  2023-03-03    n       90
## 588  2023-03-22    t      100
## 589  2023-03-12    c       50
## 590  2023-03-17    o       10
## 591  2023-03-02    c       60
## 592  2023-02-24    m      100
## 593  2023-02-27    w       90
## 594  2023-03-21    i       50
## 595  2023-03-09    y       40
## 596  2023-02-23    i       10
## 597  2023-03-22    f       90
## 598  2023-03-19    s       50
## 599  2023-03-02    n       70
## 600  2023-02-25    d       30
## 601  2023-03-24    r        0
## 602  2023-03-08    k      100
## 603  2023-03-23    b      100
## 604  2023-03-01    p       60
## 605  2023-03-01    h        0
## 606  2023-03-02    b       70
## 607  2023-03-18    z       10
## 608  2023-03-20    r       10
## 609  2023-03-23    a       80
## 610  2023-03-24    c       40
## 611  2023-02-26    r      100
## 612  2023-03-22    a       70
## 613  2023-03-23    z        0
## 614  2023-02-23    e        0
## 615  2023-03-08    z       30
## 616  2023-03-16    e        0
## 617  2023-03-11    v       80
## 618  2023-03-04    y       30
## 619  2023-03-23    m       10
## 620  2023-02-26    r       80
## 621  2023-03-17    f       90
## 622  2023-03-12    d       50
## 623  2023-03-03    r       60
## 624  2023-03-05    m       20
## 625  2023-03-08    k       40
## 626  2023-03-01    v      100
## 627  2023-03-07    r       30
## 628  2023-03-03    j       80
## 629  2023-03-02    y       10
## 630  2023-03-14    z       40
## 631  2023-02-28    g       70
## 632  2023-03-08    c       40
## 633  2023-03-21    p       90
## 634  2023-02-27    z       50
## 635  2023-03-13    j       30
## 636  2023-02-25    u       50
## 637  2023-03-17    y       60
## 638  2023-02-26    z      100
## 639  2023-03-18    x       70
## 640  2023-03-01    l       40
## 641  2023-03-05    r       10
## 642  2023-03-16    v      100
## 643  2023-02-28    u        0
## 644  2023-03-15    h       70
## 645  2023-03-09    d       20
## 646  2023-03-20    t       30
## 647  2023-03-07    m       20
## 648  2023-03-11    z       10
## 649  2023-03-05    g       90
## 650  2023-03-24    l       60
## 651  2023-03-02    b        0
## 652  2023-03-04    r       10
## 653  2023-03-08    x       80
## 654  2023-03-03    v       70
## 655  2023-03-18    v       70
## 656  2023-02-26    h       90
## 657  2023-03-21    j       60
## 658  2023-03-08    q        0
## 659  2023-03-06    w       60
## 660  2023-03-05    k       70
## 661  2023-03-11    o       80
## 662  2023-03-08    c       70
## 663  2023-03-23    z       40
## 664  2023-03-11    r       40
## 665  2023-03-02    r       80
## 666  2023-03-02    u       20
## 667  2023-03-10    r       90
## 668  2023-03-07    k       30
## 669  2023-03-15    y       40
## 670  2023-03-11    s       60
## 671  2023-03-07    f       10
## 672  2023-03-16    n       90
## 673  2023-03-20    w       70
## 674  2023-03-17    t      100
## 675  2023-02-26    v       90
## 676  2023-02-27    q       90
## 677  2023-02-26    b       20
## 678  2023-03-13    s       30
## 679  2023-03-04    r       80
## 680  2023-02-24    w       20
## 681  2023-02-28    e       70
## 682  2023-03-19    n       30
## 683  2023-03-19    e       80
## 684  2023-03-24    s       20
## 685  2023-03-11    l       40
## 686  2023-03-21    e       20
## 687  2023-02-26    s        0
## 688  2023-03-06    m       20
## 689  2023-03-05    i      100
## 690  2023-02-23    c       50
## 691  2023-03-02    n      100
## 692  2023-03-16    m       40
## 693  2023-03-05    v        0
## 694  2023-03-23    n       60
## 695  2023-03-22    a       20
## 696  2023-03-14    f       70
## 697  2023-03-15    l       50
## 698  2023-02-23    s       90
## 699  2023-03-24    x       80
## 700  2023-03-11    e       70
## 701  2023-03-12    j      100
## 702  2023-03-18    l      100
## 703  2023-03-12    r        0
## 704  2023-02-27    a       60
## 705  2023-03-13    s       20
## 706  2023-03-12    h       60
## 707  2023-03-22    e       70
## 708  2023-03-21    c       40
## 709  2023-03-05    y       70
## 710  2023-03-06    y       40
## 711  2023-02-23    x       20
## 712  2023-03-21    h       40
## 713  2023-03-09    p       60
## 714  2023-03-24    j       90
## 715  2023-03-04    h       70
## 716  2023-03-05    u       80
## 717  2023-02-26    i       20
## 718  2023-03-19    i      100
## 719  2023-02-26    g       90
## 720  2023-03-15    i        0
## 721  2023-03-11    m        0
## 722  2023-03-19    z       20
## 723  2023-03-10    t        0
## 724  2023-02-24    b       30
## 725  2023-03-17    c       70
## 726  2023-02-28    v       40
## 727  2023-03-02    x       40
## 728  2023-03-10    b       60
## 729  2023-03-07    i        0
## 730  2023-02-28    y       50
## 731  2023-03-23    q       80
## 732  2023-03-18    m       20
## 733  2023-03-24    d        0
## 734  2023-03-12    l       40
## 735  2023-03-06    g       40
## 736  2023-03-23    n       50
## 737  2023-03-16    f       30
## 738  2023-03-07    j       40
## 739  2023-03-19    k       80
## 740  2023-03-07    j       70
## 741  2023-02-27    s        0
## 742  2023-02-24    s       90
## 743  2023-02-24    u       30
## 744  2023-03-08    r       90
## 745  2023-03-16    v      100
## 746  2023-03-12    a        0
## 747  2023-02-26    o       80
## 748  2023-03-01    w       60
## 749  2023-03-06    k       60
## 750  2023-03-18    e       60
## 751  2023-03-21    l       50
## 752  2023-03-01    g        0
## 753  2023-03-03    p        0
## 754  2023-02-26    a       70
## 755  2023-03-19    a       40
## 756  2023-03-09    j       20
## 757  2023-03-14    w       70
## 758  2023-03-14    m       60
## 759  2023-03-06    r       90
## 760  2023-03-12    v       20
## 761  2023-02-23    b       70
## 762  2023-03-23    k       10
## 763  2023-02-26    a       80
## 764  2023-03-18    n       60
## 765  2023-02-26    f       30
## 766  2023-03-23    i       80
## 767  2023-03-12    p       80
## 768  2023-03-01    x       90
## 769  2023-03-06    l        0
## 770  2023-03-10    g      100
## 771  2023-03-01    t       90
## 772  2023-03-18    j      100
## 773  2023-03-14    e       60
## 774  2023-03-10    n       90
## 775  2023-03-06    f       50
## 776  2023-03-15    f       70
## 777  2023-03-19    r       70
## 778  2023-02-23    p        0
## 779  2023-03-05    g       70
## 780  2023-03-10    h       90
## 781  2023-03-19    u        0
## 782  2023-03-20    n        0
## 783  2023-03-12    q       90
## 784  2023-03-16    q       10
## 785  2023-03-13    p      100
## 786  2023-02-25    y       20
## 787  2023-03-05    r      100
## 788  2023-03-01    f        0
## 789  2023-02-23    a        0
## 790  2023-03-11    j       20
## 791  2023-03-23    k        0
## 792  2023-03-13    m       40
## 793  2023-03-08    p       70
## 794  2023-03-01    a       10
## 795  2023-03-17    c       80
## 796  2023-03-16    j       50
## 797  2023-03-12    b       30
## 798  2023-03-13    u       40
## 799  2023-03-03    r       70
## 800  2023-03-17    i       40
## 801  2023-03-20    m      100
## 802  2023-03-12    s       50
## 803  2023-03-11    l       80
## 804  2023-03-16    c       30
## 805  2023-03-15    j       40
## 806  2023-03-04    c       90
## 807  2023-03-03    a        0
## 808  2023-03-15    d       90
## 809  2023-03-06    h       20
## 810  2023-03-07    n       10
## 811  2023-03-11    j       90
## 812  2023-02-26    r       50
## 813  2023-02-28    w       40
## 814  2023-03-10    b       40
## 815  2023-02-25    s       60
## 816  2023-03-15    f       40
## 817  2023-03-23    f       40
## 818  2023-03-07    f       50
## 819  2023-03-17    k       10
## 820  2023-03-09    y       10
## 821  2023-03-20    r       30
## 822  2023-03-05    w      100
## 823  2023-03-01    y      100
## 824  2023-02-23    o       80
## 825  2023-03-02    c       90
## 826  2023-03-18    u       70
## 827  2023-03-22    y       50
## 828  2023-03-16    s       50
## 829  2023-03-24    r       40
## 830  2023-03-21    r       80
## 831  2023-03-10    l       40
## 832  2023-03-09    t       20
## 833  2023-03-23    b       20
## 834  2023-03-11    k       40
## 835  2023-03-24    a       70
## 836  2023-03-06    g       90
## 837  2023-03-23    r       70
## 838  2023-03-06    x       10
## 839  2023-03-20    k      100
## 840  2023-02-23    i       40
## 841  2023-03-14    f       30
## 842  2023-02-23    c       20
## 843  2023-03-07    e       40
## 844  2023-03-19    c       50
## 845  2023-03-05    p        0
## 846  2023-03-17    r       40
## 847  2023-02-26    r        0
## 848  2023-03-23    d       60
## 849  2023-03-20    f       90
## 850  2023-03-02    b       60
## 851  2023-03-10    n       50
## 852  2023-03-02    o       50
## 853  2023-03-15    u       50
## 854  2023-03-21    q      100
## 855  2023-03-17    t       40
## 856  2023-02-24    n        0
## 857  2023-02-28    m       90
## 858  2023-03-18    n       20
## 859  2023-03-18    b       50
## 860  2023-03-07    n       10
## 861  2023-02-28    c       60
## 862  2023-02-24    c       40
## 863  2023-03-16    n       60
## 864  2023-03-20    z       50
## 865  2023-03-23    y       50
## 866  2023-03-11    q       40
## 867  2023-03-14    n       40
## 868  2023-03-03    p       40
## 869  2023-03-15    b       60
## 870  2023-03-20    j       80
## 871  2023-03-06    i       10
## 872  2023-03-07    m       80
## 873  2023-03-22    w       60
## 874  2023-03-15    s       10
## 875  2023-03-13    p       90
## 876  2023-03-12    h       90
## 877  2023-03-09    k        0
## 878  2023-03-10    t       30
## 879  2023-03-18    b      100
## 880  2023-03-15    f       30
## 881  2023-03-22    w       70
## 882  2023-03-06    m       70
## 883  2023-03-07    f       30
## 884  2023-03-05    d       40
## 885  2023-02-26    p        0
## 886  2023-03-11    a      100
## 887  2023-03-16    o      100
## 888  2023-02-27    u       10
## 889  2023-02-24    b       50
## 890  2023-03-10    g       10
## 891  2023-03-17    i       20
## 892  2023-03-05    c       20
## 893  2023-02-25    a       50
## 894  2023-03-14    v       10
## 895  2023-03-06    h       20
## 896  2023-03-23    e       10
## 897  2023-03-08    p       50
## 898  2023-03-14    q       40
## 899  2023-02-24    j       70
## 900  2023-03-18    d       50
## 901  2023-02-26    z      100
## 902  2023-03-07    n      100
## 903  2023-03-22    b       80
## 904  2023-03-16    r       80
## 905  2023-03-19    t       30
## 906  2023-03-15    d       80
## 907  2023-03-18    f       80
## 908  2023-02-25    o       10
## 909  2023-03-15    y       50
## 910  2023-02-28    p       40
## 911  2023-03-01    j       60
## 912  2023-03-07    e        0
## 913  2023-02-28    v       90
## 914  2023-02-24    t       70
## 915  2023-03-17    q      100
## 916  2023-02-25    o        0
## 917  2023-03-09    u        0
## 918  2023-03-22    z      100
## 919  2023-03-21    g      100
## 920  2023-03-06    m       20
## 921  2023-02-27    j       50
## 922  2023-03-21    p        0
## 923  2023-03-14    i       60
## 924  2023-03-22    o       30
## 925  2023-03-22    b       30
## 926  2023-03-23    m       90
## 927  2023-03-03    e       50
## 928  2023-03-12    o        0
## 929  2023-03-16    j       80
## 930  2023-02-23    j       50
## 931  2023-03-04    a       80
## 932  2023-03-11    t       90
## 933  2023-03-10    a       40
## 934  2023-02-28    f       90
## 935  2023-03-06    u       20
## 936  2023-03-15    o       40
## 937  2023-03-03    v       10
## 938  2023-03-04    c       30
## 939  2023-03-11    a       20
## 940  2023-02-24    z       50
## 941  2023-03-10    s       50
## 942  2023-03-06    s       40
## 943  2023-03-16    r       70
## 944  2023-03-15    j       70
## 945  2023-02-27    r       80
## 946  2023-03-10    h       10
## 947  2023-03-07    x      100
## 948  2023-03-16    t       80
## 949  2023-03-23    x       90
## 950  2023-03-03    t       50
## 951  2023-03-01    q       90
## 952  2023-03-05    k       50
## 953  2023-03-18    o       70
## 954  2023-03-03    u       60
## 955  2023-03-16    h       40
## 956  2023-03-19    n        0
## 957  2023-03-15    x       90
## 958  2023-03-03    i       20
## 959  2023-03-08    d       90
## 960  2023-03-11    z       50
## 961  2023-03-08    b       10
## 962  2023-03-05    l       50
## 963  2023-03-12    q        0
## 964  2023-03-03    c       30
## 965  2023-03-23    w        0
## 966  2023-03-10    w       30
## 967  2023-03-14    x       60
## 968  2023-03-22    n       60
## 969  2023-03-17    h       30
## 970  2023-03-13    h        0
## 971  2023-02-25    x       30
## 972  2023-03-16    d       50
## 973  2023-03-04    w       30
## 974  2023-03-06    k        0
## 975  2023-03-19    g       50
## 976  2023-03-11    o       10
## 977  2023-03-17    p       80
## 978  2023-02-27    q       60
## 979  2023-03-24    g        0
## 980  2023-03-15    o        0
## 981  2023-02-23    h       30
## 982  2023-03-12    o       70
## 983  2023-03-20    e       80
## 984  2023-02-26    z        0
## 985  2023-03-17    h      100
## 986  2023-03-08    k       60
## 987  2023-03-07    q       40
## 988  2023-03-09    g       60
## 989  2023-02-23    h       10
## 990  2023-03-23    u       20
## 991  2023-03-20    f       50
## 992  2023-03-13    g       80
## 993  2023-03-02    i        0
## 994  2023-02-28    u       70
## 995  2023-03-24    v       40
## 996  2023-03-10    b       80
## 997  2023-03-19    j      100
## 998  2023-03-13    x      100
## 999  2023-03-01    p       30
## 1000 2023-03-15    l       90
```

---

- Warns you when you make mistakes (!):

```r
tib1$quanitty
```

<pre style="color: red;"><code>## Warning: Unknown or uninitialised column: `quanitty`.
</code></pre>

```
## NULL
```

```r
df1$quanitty
```

```
## NULL
```

---

- Can also create via `tribble()`:

```r
tribble(
  ~a, ~b, ~c,
  "a", 1, 2.2,
  "b", 2, 4.3,
  "c", 3, 3.4
)
```

```
## # A tibble: 3 × 3
##   a         b     c
##   <chr> <dbl> <dbl>
## 1 a         1   2.2
## 2 b         2   4.3
## 3 c         3   3.4
```

---

- Can build on top of variables during creation:

```r
tibble(x = 1:5, y = x^2)
```

```
## # A tibble: 5 × 2
##       x     y
##   <int> <dbl>
## 1     1     1
## 2     2     4
## 3     3     9
## 4     4    16
## 5     5    25
```

```r
data.frame(x = 1:5, y = x^2)
```

<pre style="color: red;"><code>## Error in data.frame(x = 1:5, y = x^2): object 'x' not found
</code></pre>

---

- Will never turn your strings into factors, will never change your column names:

```r
tib1 <- readr::read_csv("../data/another_bad_test_args.csv", col_types = cols())
colnames(tib1)
```

```
## [1] "col 1" "col 2" "col 3"
```

```r
tib1$`col 3`
```

```
## [1]  2  3 NA  4
```

```r
df1 <- read.csv("../data/another_bad_test_args.csv")
colnames(df1)
```

```
## [1] "col.1" "col.2" "col.3"
```

```r
df1$col.3
```

```
## [1]  2  3 NA  4
```

---

Though one ought to remember a `tibble` is still a `data.frame`:

```r
class(tib1)
```

```
## [1] "spec_tbl_df" "tbl_df"      "tbl"         "data.frame"
```

```r
class(df1)
```

```
## [1] "data.frame"
```

---
class: section-slide

# End of Detour

---

## Consistency - Example: `stringr`

> a cohesive set of functions designed to make working with strings as easy as possible.

```r
strings_vec <- c("I'm feeling fine", "I'm perfectly OK",
                 "Nothing is wrong!")
str_length(strings_vec)
```

```
## [1] 16 16 17
```

```r
str_c(strings_vec, collapse = ", ")
```

```
## [1] "I'm feeling fine, I'm perfectly OK, Nothing is wrong!"
```

```r
str_sub(strings_vec, 1, 3)
```

```
## [1] "I'm" "I'm" "Not"
```

---

```r
str_detect(strings_vec, "I'm")
```

```
## [1]  TRUE  TRUE FALSE
```

```r
str_replace(strings_vec, "I'm", "You're")
```

```
## [1] "You're feeling fine" "You're perfectly OK" "Nothing is wrong!"
```

```r
str_split("Do you know regex?", " ")
```

```
## [[1]]
## [1] "Do"     "you"    "know"   "regex?"
```

```r
str_extract(strings_vec, "[aeiou]")
```

```
## [1] "e" "e" "o"
```

```r
str_count(strings_vec, "[A-Z]")
```

```
## [1] 1 3 1
```

---

## The Pipe

Remember you?

data.frame(orientation = c("bisexual", "gay", "straight"),
           income_mean = c(mean_bi, mean_gay, mean_straight))
```

```
##   orientation income_mean
## 1    bisexual   133421.05
## 2         gay    86489.36
## 3    straight    85219.74
```

---

Doesn't this make much more sense?

```r
okcupid %>%
  filter(sex == "f", age > 30) %>%
  group_by(orientation) %>%
  summarize(income_mean = mean(income, na.rm = TRUE))
```

```
## # A tibble: 3 × 2
##   orientation income_mean
##   <chr>             <dbl>
## 1 bisexual        133421.
## 2 gay              86489.
## 3 straight         85220.
```

- Read as:
  - Take the OKCupid data,
  - Filter only women above the age of 30,
  - And for each group of sexual orientation,
  - Give me the average income

---

- Make verbs, not nouns
- Can always access the dataset last stage with "`.`":

```r
okcupid %>%
  filter(str_count(essay0) > median(str_count(.$essay0), na.rm = T))
```
- Operates not just on data frames or tibbles:

```r
strings_vec %>% str_to_title()
```

```
## [1] "I'm Feeling Fine"  "I'm Perfectly Ok"  "Nothing Is Wrong!"
```
- No intermediate objects
- Don't strive to make the longest possible pipe (though it is a fun experiment)
- Tools exist for debugging (e.g. `%T>%`, the ViewPipeStep package, ...)

---

And, if you want to throw in the n, the median:

```r
okcupid %>%
  filter(sex == "f", age > 30) %>%
  group_by(orientation) %>%
  summarize(income_mean = mean(income, na.rm = TRUE),
            income_median = median(income, na.rm = TRUE),
            n = n())
```

```
## # A tibble: 3 × 4
##   orientation income_mean income_median     n
##   <chr>             <dbl>         <dbl> <int>
## 1 bisexual        133421.         50000   652
## 2 gay              86489.         40000   664
## 3 straight         85220.         60000 10436
```

---

And if you want this for the age as well:

```r
okcupid %>%
  filter(sex == "f", age > 30) %>%
  group_by(orientation) %>%
  summarize(across(c(income, age),
                   list(mean = ~mean(.x, na.rm = TRUE),
                        median = ~median(.x, na.rm = TRUE))))
```

```
## # A tibble: 3 × 5
##   orientation income_mean income_median age_mean age_median
##   <chr>             <dbl>         <dbl>    <dbl>      <dbl>
## 1 bisexual        133421.         50000     37.8         36
## 2 gay              86489.         40000     40.4         38
## 3 straight         85220.         60000     40.7         38
```

Now *this* is a language for Data Science.

But we're getting ahead of ourselves.

---

### In fact!

This became so popular that since R 4.1 there is a built-in pipe operator:

```r
okcupid |>
  filter(sex == "f", age > 30) |>
  group_by(orientation) |>
  summarize(across(c(income, age),
            list(mean = \(x) mean(x, na.rm = TRUE),
                 median = \(x) median(x, na.rm = TRUE))))
```

---

## `ggplot2`

.font80percent[
[Ave Mariah / Giora Simchoni](http://giorasimchoni.com/2017/12/10/2017-12-10-ave-mariah/)
]
---

## `ggplot2`

.font80percent[
[Soviet Space Dogs / David Smale](https://davidsmale.netlify.com/portfolio/soviet-space-dogs-part-2/)
]
---

## `ggplot2`

.font80percent[
[Federer, Nadal, Djokovic and Murray, Love. / Giora Simchoni](http://giorasimchoni.com/2017/05/01/2017-05-01-federer-nadal-djokovic-and-murray-love/)
]
---

## `ggplot2`

.font80percent[
[NYT-style urban heat island maps / Katie Jolly](https://www.katiejolly.io/blog/2019-08-28/nyt-urban-heat)
]
---

## `ggplot2`

.font80percent[
[A map of marriage rates, state by state / Unkown](https://www.r-graph-gallery.com/328-hexbin-map-of-the-usa.html)
]
---

## `ggplot2`

.font80percent[
[Calendar-based graphics for visualizing people’s daily schedules / Earo Wang](https://pdf.earo.me/calendar-vis.pdf)
]
---

## The Community

.pull-left[
- 100% Open Source on Github
- Cheatsheet for everything
- Documentation for humans, Packages websites, Webinars, Free Books (start with [R4DS](https://r4ds.had.co.nz/))
- [Rstudio Community forum](https://community.rstudio.com/)
- [RLadies](https://rladies.org/) worldwide branches .font80percent[(who will pick up the 🥊 and create RLadies TLV?)]
- Very strong on Twitter [#rstats](https://twitter.com/search?q=%23rstats)
]

.pull-right[
<a href="https://rstudio.com/resources/cheatsheets/"><img src="images/stringr_cheatsheet.png" style="width: 100%" /></a>
]