eda_paged_report: Reporting the information of EDA

Description

The eda_paged_report() paged report the information for EDA.

Usage

eda_paged_report(.data, ...)
# S3 method for data.frame
eda_paged_report(
  .data,
  target = NULL,
  output_format = c("pdf", "html"),
  output_file = NULL,
  output_dir = tempdir(),
  browse = TRUE,
  title = "EDA Report",
  subtitle = deparse(substitute(.data)),
  author = "dlookr",
  abstract_title = "Report Overview",
  abstract = NULL,
  title_color = "black",
  subtitle_color = "blue",
  cover_img = NULL,
  create_date = Sys.time(),
  logo_img = NULL,
  theme = c("orange", "blue"),
  sample_percent = 100,
  is_tbl_dbi = FALSE,
  ...
)

Arguments

.data

a data.frame or a tbl_df.

...

arguments to be passed to methods.

target

character. target variable.

output_format

report output type. Choose either "pdf" and "html". "pdf" create pdf file by rmarkdown::render() and pagedown::chrome_print(). so, you needed Chrome web browser on computer. "html" create html file by rmarkdown::render().

output_file

name of generated file. default is NULL.

output_dir

name of directory to generate report file. default is tempdir().

browse

logical. choose whether to output the report results to the browser.

title

character. title of report. default is "Data Diagnosis Report".

subtitle

character. subtitle of report. default is name of data.

author

character. author of report. default is "dlookr".

abstract_title

character. abstract title of report. default is "Report Overview".

abstract

character. abstract of report.

title_color

character. color of title. default is "black".

subtitle_color

character. color of subtitle. default is "blue".

cover_img

character. name of cover image.

create_date

Date or POSIXct, character. The date on which the report is generated. The default value is the result of Sys.time().

logo_img

character. name of logo image file on top right.

theme

character. name of theme for report. support "orange" and "blue". default is "orange".

sample_percent

numeric. Sample percent of data for performing Diagnosis. It has a value between (0, 100]. 100 means all data, and 5 means 5% of sample data. This is useful for data with a large number of observations.

is_tbl_dbi

logical. whether .data is a tbl_dbi object.

Reported information

The EDA process will report the following information:

Overview
- Data Structures
- Job Informations
Univariate Analysis
- Descriptive Statistics
  - Numerical Variables
  - Categorical Variables
- Normality Test
Bivariate Analysis
- Compare Numerical Variables
- Compare Categorical Variables
Multivariate Analysis
- Correlation Analysis
  - Correlation Coefficient Matrix
  - Correlation Plot
Target based Analysis
- Grouped Numerical Variables
- Grouped Categorical Variables
- Grouped Correlation

Details

Generate generalized EDA report automatically. You can choose to output to pdf and html files. This feature is useful for EDA of data with many variables, rather than data with fewer variables.

Create an PDF through the Chrome DevTools Protocol. If you want to create PDF, Google Chrome or Microsoft Edge (or Chromium on Linux) must be installed prior to using this function. If not installed, you must use output_format = "html".

Examples

Run this code

# NOT RUN {
# create the dataset
heartfailure2 <- dlookr::heartfailure
heartfailure2[sample(seq(NROW(heartfailure2)), 20), "sodium"] <- NA
heartfailure2[sample(seq(NROW(heartfailure2)), 5), "smoking"] <- NA

# create pdf file. file name is EDA_Paged_Report.pdf
# eda_paged_report(heartfailure2, sample_percent = 80)

# create pdf file. file name is EDA.pdf. and change cover image
# cover <- file.path(system.file(package = "dlookr"), "report", "cover1.jpg")
# eda_paged_report(heartfailure2, cover_img = cover, title_color = "gray",
#   output_file = "EDA.pdf")

# create pdf file. file name is ./EDA.pdf and not browse
# cover <- file.path(system.file(package = "dlookr"), "report", "cover3.jpg")
# eda_paged_report(heartfailure2, output_dir = ".", cover_img = cover, 
#   flag_content_missing = FALSE, output_file = "EDA.pdf", browse = FALSE)

# create pdf file. file name is EDA_Paged_Report.html
# eda_paged_report(heartfailure2, target = "death_event", output_format = "html")
# }
# NOT RUN {
# }