README.Rmd

---
output:
  github_document
---


```{r, echo = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE, 
  comment = "#>",
  fig.path = "README-"  
)
options(width = 110)
```


### Installation

You can install:

* the latest development version from GitHub with

    ```R
    install.packages("devtools")
    devtools::install_github("ryanburge/blocks/building")
    ```
    
### There are just a handful of functions to building right now 


## Counting Things

I love the functionality of tabyl, but it doesn't take a weight variable. Here's the simple version `ct()`

```{r, warning = FALSE, message = FALSE}
library(tidyverse)
library(building)
cces <- read_csv("https://raw.githubusercontent.com/ryanburge/blocks/master/cces.csv")

cces %>% 
  ct(race)
```

Note that you are presented with a count column and a pct column. 

Let's add weights

```{r, warning = FALSE, message = FALSE}
cces %>% 
  ct(race, commonweight_vv)
```

Notice that it's pipeable. And if you don't include the weight variable then it won't be calculated with a weight. 


## Getting Confidence Intervals

Oftentimes in social science we like to see what our 95% confidence intervals are, but that's a lot of syntax. It's easy with the `mean_ci` function

```{r, warning = FALSE, message = FALSE}
cces %>% 
  mean_ci(gender)
```

The default is a 95% confidence interval. However that can be changed easily. 


```{r, warning = FALSE, message = FALSE}
cces %>% 
  mean_ci(gender, ci = .84)
```


## Simple Mean and Median 

I wanted a simple function to calculate the mean and the median. It takes just one variable and computes both statistics. 

```{r, warning = FALSE, message = FALSE}
money1 <- read_csv("https://raw.githubusercontent.com/ryanburge/pls2003_sp17/master/sal_work.csv")

money1 

money1  %>% 
  mean_med(salary)
```


## Two Value Correlations 

Here's a simple function that generates a pearson correlation of two variables with a p-value.  

```{r, warning = FALSE, message = FALSE}
x <- c(1, 2, 3, 7, 5, 777, 6, 411, 8)
y <- c(11, 23, 1, 4, 6, 22455, 34, 22, 22)
z <- c(34, 3, 21, 4555, 75, 2, 3334, 1122, 22312)

test <- data.frame(x,y,z) %>% as.tibble()

test %>% 
  filter(z > 10) %>% 
  corr(x,y)

```


## Convert all NAs in a dataframe to Zero

```{r, warning = FALSE, message = FALSE}
x <- c(1, 2, 3, NA, 5, NA, 6, NA, 8)
y <- c(11, 23, NA, 4, 6, NA, NA, NA, 22)

df <- data.frame(x,y) %>% as.tibble()

df

na_zero(df, y)

```

## Bind Several Dataframes together 

Oftentimes I make many little dataframes that I need to bind_rows to put into one large dataframe. As long as those dataframes have the same naming convention that can be done. 

```{r, warning = FALSE, message = FALSE}
dd1 <- data.frame(a = 1, b = 2)
dd2 <- data.frame(a = 3, b = 4)
dd3 <- data.frame(a = 5, b = 6) 

bind_df("dd")
```




## Recode things and keep the factor levels 

I recode all the time, but unfortunately when you recode from numeric to character the factor levels are plotted in alphabetical order. There's a way around that now. 

I found this [terrific function](https://stackoverflow.com/questions/49572416/r-convert-to-factor-with-order-of-levels-same-with-case-when) written by [Dennis YL](https://stackoverflow.com/users/5068121/dennis-yl), where he had the same problem that I had. 

```{r, warning = FALSE, message = FALSE, fig.width= 12}
cces <- read_csv("https://raw.githubusercontent.com/ryanburge/cces/master/CCES%20for%20Methods/small_cces.csv")

graph <- cces %>% 
  mutate(pid_new = frcode(pid7 == 1 ~ "Strong Democrat", 
                          pid7 == 2 ~ "Not Strong Democrat", 
                          pid7 == 3 ~ "Lean Democrat", 
                          pid7 == 4 ~ "Independent", 
                          pid7 == 5 ~ "Lean Republican", 
                          pid7 == 6 ~ "Not Strong Republican", 
                          pid7 == 7 ~ "Strong Republican", 
                          TRUE ~ "REMOVE")) %>% 
  ct(pid_new)


graph %>% 
  filter(pid_new != "REMOVE") %>% 
  ggplot(., aes(x = pid_new, y = pct)) +
  geom_col()
```

* let me know what you think on twitter <a href="https://twitter.com/ryanburge">@ryanburge</a>