https://github.com/data-cleaning/validatedb

Validate on a table in a DB, using dbplyr
https://github.com/data-cleaning/validatedb

database datacleaning validation

Last synced: 3 months ago
JSON representation

Validate on a table in a DB, using dbplyr

• Host: GitHub
• URL: https://github.com/data-cleaning/validatedb
• Owner: data-cleaning
• License: gpl-3.0
• Created: 2020-11-11T23:18:56.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2022-06-03T14:11:33.000Z (over 2 years ago)
• Last Synced: 2024-06-05T02:32:08.925Z (9 months ago)
• Topics: database, datacleaning, validation
• Language: R
• Homepage: https://data-cleaning.github.io/validatedb
• Size: 623 KB
• Stars: 32
• Watchers: 2
• Forks: 4
• Open Issues: 4
• Metadata Files:
  • Readme: README.Rmd
  • License: LICENSE.md

Awesome Lists containing this project

• jimsghstars - data-cleaning/validatedb - Validate on a table in a DB, using dbplyr (R)

README

        
---

output: github_document

---

```{r, include = FALSE}

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%"

)

```

  

# validatedb

[![CRAN status](https://www.r-pkg.org/badges/version/validatedb)](https://CRAN.R-project.org/package=validatedb)

[![R build status](https://github.com/data-cleaning/validatedb/workflows/R-CMD-check/badge.svg)](https://github.com/data-cleaning/validatedb/actions)

[![Codecov test

coverage](https://codecov.io/gh/data-cleaning/validatedb/branch/master/graph/badge.svg)](https://codecov.io/gh/data-cleaning/validatedb?branch=master)

[![Mentioned in Awesome Official Statistics ](https://awesome.re/mentioned-badge.svg)](http://www.awesomeofficialstatistics.org)

`validatedb` executes validation checks written with R package `validate` on a 

database. This allows for checking the validity of records in a database.

## Installation

You can install a development version with 

``` r

remotes::install_github("data-cleaning/validatedb")

```

## Example

```{r example}

library(validatedb)

```

First we setup a table in a database (for demo purpose)

```{r}

# create a table in a database

income <- data.frame(id=1:2, age=c(12,35), salary = c(1000,NA))

con <- DBI::dbConnect(RSQLite::SQLite())

DBI::dbWriteTable(con, "income", income)

```

We retrieve a reference/handle to the table in the DB with `dplyr`

```{r}

tbl_income <- tbl(con, "income")

print(tbl_income)

```

Let's define a rule set and confront the table with it:

```{r}

rules <- validator( is_adult   = age >= 18

                  , has_income = salary > 0

                  , mean_age   = mean(age,na.rm=TRUE) > 24

                  , has_values = is_complete(age, salary)

                  )

# and confront!

cf <- confront(tbl_income, rules, key = "id")

print(cf)

summary(cf)

```

Values (i.e. validations on the table) can be retrieved like in `validate` with

`type="matrix"` or `type="list"`

```{r}

values(cf, type = "matrix")

```

But often this seems more handy:

```{r}

values(cf, type = "tbl")

```

or 

```{r}

values(cf, type = "tbl", sparse=TRUE)

```

We can see the sql code by using `show_query`:

```{r}

show_query(cf)

```

Or write the sql to a file for documentation (and inspiration)

```{r, eval = FALSE}

dump_sql(cf, "validation.sql")

```

```sql

```{r, echo = FALSE, results="asis"}

dump_sql(cf)

```

```

### Aggregate example

```{r aggregate, code = readLines("./example/aggregate.R")}

```

## validate specific functions

### Added:

- [x] `is_complete`, `all_complete`

- [x] `is_unique`, `all_unique`

- [x] `exists_any`, `exists_one`

- [x] `do_by`, `sum_by`, `mean_by`, `min_by`, `max_by` 

### Todo

Some newly added `validate` utility functions are (still) missing from `validatedb`.

- [ ] `contains_exactly`

- [ ] `is_linear_sequence`

- [ ] `hierachy`