boxuancui / dataexplorer Goto Github PK

Reference: http://stackoverflow.com/questions/25819539/how-to-add-an-interactive-visualization-to-r-markdown

Thanks for the library!

Would it be possible to add an option for a title to the plots?

I think all it would require is
function(data, title=NULL) {
[ggplot code] +
ggtitle(title)

Ryan

1. All discrete, no continuous.
2. All continuous, no discrete.
3. 1 discrete, multiple continuous.
4. 1 continuous, multiple discrete.
5. 1 continuous with mostly NA.

Hey sir! Wondering if it's possible to set it up so the library installs rprojroot as part of the install process. I didn't have it installed and as such it popped an error for me.

1. Add Travis CI for branch "develop"
2. Add badge: http://www.r-pkg.org/badges/version/rpackage

Warning message from data.table:

Warning message:
In [.data.table(data, , :=(ind, NULL), with = FALSE) :
with=FALSE together with := was deprecated in v1.9.4 released Oct 2014. Please wrap the LHS of := with parentheses; e.g., DT[,(myVar):=sum(b),by=a] to assign to column name(s) held in variable myVar. See ?':=' for other examples. As warned in 2014, this is now a warning.

Add SetNaTo().

If selected feature is discrete, do box plot for all categories. Otherwise, split continuous features into quartiles and treat it as discrete.

Do not use cat().

checking Rd cross-references ... WARNING
Missing link or links in documentation object 'BarDiscrete.Rd':
'data.table'
Missing link or links in documentation object 'CollapseCategory.Rd':
'data.table'
Missing link or links in documentation object 'CorrelationContinuous.Rd':
'data.table'
Missing link or links in documentation object 'CorrelationDiscrete.Rd':
'data.table'
Missing link or links in documentation object 'DensityContinuous.Rd':
'data.table'
Missing link or links in documentation object 'DropVar.Rd':
'data.table'
Missing link or links in documentation object 'GenerateReport.Rd':
'data.table'
Missing link or links in documentation object 'HistogramContinuous.Rd':
'data.table'
Missing link or links in documentation object 'PlotMissing.Rd':
'data.table'
Missing link or links in documentation object 'SetNaTo.Rd':
'data.table'
Missing link or links in documentation object 'SplitColType.Rd':
'data.table'
See section 'Cross-references' in the 'Writing R Extensions' manual.

https://www.r-project.org/nosvn/R.check/r-devel-windows-ix86+x86_64/DataExplorer-00check.html

• Add ability to collapse based on desired columns.
• Add ability to collapse based on multiple columns.

When quiet argument is not supplied, the function will not print report location, instead, generate an error.

num_all_missing is missing.

In addition,

1. exclude all-missing continuous features.
2. fix wrong input to argument binwidth causing error.

Hi,

I've just tried DataExplorer and have followed the minimal example:

library(DataExplorer)
GenerateReport(iris)

It seems that the knitr part runs OK, but then it stops at the end with error:

Error in file(con, "w") : cannot open the connection
In addition: Warning message:
In file(con, "w") :  cannot open file 'report.knit.md': Permission denied

I have also tried the diamonds example with the same result.

Any hint on what may happen?

Thank you,

> sessionInfo()
R version 3.2.4 (2016-03-10)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Gentoo/Linux

locale:
 [1] LC_CTYPE=ca_AD.UTF8        LC_NUMERIC=C               LC_TIME=ca_AD.UTF8         LC_COLLATE=C               LC_MONETARY=ca_AD.UTF8     LC_MESSAGES=ca_AD.UTF8     LC_PAPER=ca_AD.UTF8
 [8] LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C             LC_MEASUREMENT=ca_AD.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] ggplot2_2.1.0      DataExplorer_0.2.4 vimcom_1.2-6       setwidth_1.0-4     colorout_1.0-3

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.3      knitr_1.12.3     magrittr_1.5     munsell_0.4.3    colorspace_1.2-6 stringr_1.0.0    plyr_1.8.3       tools_3.2.4      grid_3.2.4       data.table_1.9.6 gtable_0.2.0     htmltools_0.3.5
[13] yaml_2.1.13      digest_0.6.9     gridExtra_2.2.1  reshape2_1.4.1   formatR_1.3      evaluate_0.8.3   rmarkdown_0.9.5  labeling_0.3     stringi_1.0-1    scales_0.4.0     chron_2.3-47
>

Worked beautifully on the sample data sets. I tried on a personal data set and received the following error:

  |......................................                           |  59%
label: density_continuous
  |..........................................                       |  65%
  ordinary text without R code

  |..............................................                   |  71%
label: correlation_continuous
Quitting from lines 51-52 (report.rmd) 
Error: Aesthetics must be either length 1 or the same as the data (100): x, y, fill

All data are transformed into data.table for speed. Might want to detect the input format and output the same format.

1. CollapseCategory
2. SplitColType
3. PlotMissing maybe?

Inspiration:

• Understanding Principal Component Analysis Using Stack Overflow Data
• Health Care Indicators In Utah Counties

In the following line, observations are not rows.

There are **`r format(nrow(data), big.mark = ",")`** observations (rows)

In addition,

1. include more information in Basic Statistics section.
2. overall organization of the report.

• Pick what to appear in report.
• Able to customize each chart.

At first: I like your package very much, it works perfect for data.frames.

But I am looking for a functionality to get quick information on more complex Data, as structured lists or multidimensional arrays. What are the names, where are the tables etc.
Or maybe, to report the content of a *.RData file (Data, functions, ...)?

Is it possible to include this in your Package?

The example looks very nice. I am trying to run it on my own data, but I am getting the following error:

label: correlation_continuous
Quitting from lines 51-52 (report.rmd) 
Error in seq.default(from = best$lmin, to = best$lmax, by = best$lstep) : 
  'from' must be of length 1

Reported by @djhurio in #16

I have installed development version. I am not getting errors any more. Report is generated with a warning:

Warning message:
In writeLines(if (encoding == "") res else native_encode(res, to = encoding),  :
  invalid char string in output conversion`

And report is unreadable.

Reported by Uros Godnov:

I've tried your package and when running GenerateReport(mydata) I get the following error:

Quitting from lines 58-59 (report.rmd)
Error in BarDiscrete(data) : No Discrete Features

CRAN check results: https://cran.r-project.org/web/checks/check_results_DataExplorer.html

Hi
Where is package "eda" that GenerateReport requires?

GenerateReport<-
function (input_data, output_file = "report.html", output_dir = getwd(),
...)
{
report_dir <- system.file("rmd_template/report.rmd", package = "eda")

Set percentage to 1 - current threshold

Something like ggplot2::guides(fill = guide_legend(nrow = 1))

To re-produce:

data <- data.frame("a" = as.factor(round(rnorm(500, 10, 5))), "b" = rexp(500, 1:500))
table(data$a)
CollapseCategory(data, "a", 0.2, update = TRUE) ## data is not updated
table(data$a)

Instead, use reshape2::dcast for all levels of a discrete feature.

Ignore and keep indicated categories when collapsing.

Reported by @Raelili :

The color scale for correlation heat map looks misleading sometimes. For example, a -0.8 correlation looks less correlated than a 0.1 correlation. The color scale should be fixed to either [-1, 1] or the maximum of both absolute values.

The feature should mostly be response variable, so should support both discrete and continuous scale.

http://style.tidyverse.org/
To deprecate old functions, see https://stackoverflow.com/a/10145627/2158269.