I'm sailing the data seas ⛵. My home ports are Greifswald, Münster, Mainz, and the Datavana.
strohne / volker Goto Github PK
View Code? Open in Web Editor NEWVolker Package
Home Page: https://strohne.github.io/volker/
License: Other
Volker Package
Home Page: https://strohne.github.io/volker/
License: Other
Can we print interactive tables to switch between column, row or total shares?
Fix all RCheck issues
Implement table titles and a base row
Currently, some functions (for counts) provide a missings parameter. Others just skip missings. Suggestion:
Implement report_metrics and report_counts
Links to vignettes in the readme do not work: https://strohne.github.io/volker/index.html
Using vignette("introduction")
or similar might help. See https://pkgdown.r-lib.org/articles/linking.html
Numbers for metric plots: Output the mean?
Some already come with ci, add for all
Use cli package to generate nice colorful messages in the check functions.
Prüfen
Show RMarkdowns with binder or https://binderhub.uni-muenster.de/
In test-labels.R, consider testing the data frames, not the printing result.
Winbuilder says:
* checking CRAN incoming feasibility ... [12s] NOTE
...
Version contains large components
When building a source package, the rendered markdowns include big plots.
Default parameter in descriptive tables should be: no missings. Can the number of missings be printed below the table?
In plot_counts_one(), plot_counts_one_grouped() and plots_counts_items() we have a category parameter. Add the same parameter to the corresponding table functions. And allow cumulative counts for ordered values.
#' TODO: Reorder boolean categories: first TRUE, then FALSE
#' TODO: Support single columns
Due to a bug in the function data_rm_negatives
, not all negative values were removed in the report, tab, plot, and effect functions.
The bug is fixed, but not yet deployed. As a workaround, install the latest development version or explicitly remove specific residual values.
The following option, if set before the original dataframe is converted to a vlkr dataframe, can be used as a workaround to remove residual values:
options(vlkr.na.numbers=c(-1,-2,-9))
For all plot / count / effects functions, check whether the scale of the input vars is as expected (e.g. metric for a metric table) and print appropriate warnings.
Check whether a column contains only NA values and output a warning (by now, it fails with cryptic messages)
In data_rm_zeros, all NAs are dropped. Limit to zeroed NAs.
Same for data_rm_negatives.
Tutorial: show how to quickly analyze Sosci Survey data
In https://strohne.github.io/volker/articles/introduction.html the codebook() function output is ugly:
Plot height is only calculated when using the report function. Implement a plot function for vlkr_plots that calculate and render the target height for other functions.
Remove plot / report tests from build?
Or tweak settings: rstudio/rstudio#7028
Support custom fill colors for ordered bar plots (VLKR_FILLGRADIENT).
For unordered:
theme_set(theme_vlkr(base_fill = list(
c("red"),
c("red","blue","green","magenta","black")))
)
Implement a parameter base_gradient.
See "Replace item labels" test in test-labels.R
Add examples to the function help
Not yet implemented. The future will come.
Order function help and add headers
First release:
usethis::use_cran_comments()
Title:
and Description:
@return
and @examples
Authors@R:
includes a copyright holder (role 'cph')Prepare for release:
git pull
urlchecker::url_check()
devtools::build_readme()
devtools::check(remote = TRUE, manual = TRUE)
devtools::check_win_devel()
git push
Submit to CRAN:
usethis::use_version('major')
devtools::submit_cran()
Wait for CRAN...
usethis::use_github_release()
usethis::use_dev_version(push = TRUE)
For all effects_* functions, not for the parameter name
In plot_counts_one_grouped, flip the plot when prop is changed to cols.
Add a simple getting started section to the readme
In all plots and tables, keep item, value and factor order.
For the readme: Create better tables and maybe plots, revise the layout. Can be directly created inside the README.Rmd
See .plot_bars
Die Reihenfolge der Items kommt durcheinander:
ds <- tribble(
~f1, ~f2, ~f10,
1, 1, 1,
1, 2, 1,
2, 2, 2
)
plot_counts(ds, c(f1:f10))
tab_counts(ds, c(f1:f10))
Bei metrics klappt es:
plot_metrics(ds, c(f1:f10))
tab_metrics(ds, c(f1:f10))
Außerdem ist hier das Farbschema unschön. Besser differenzieren:
We have a limits parameter in metric plot function. It does nothin yet.
In zip_tables implement line breaks as separator
Implement category-parameter in plot_counts_one() as in all other plots methods for categorical data
In frequency plots, the value order may be changed (-1, +1). Keep the residual categories with negative values always as last category.
Prepare for release:
git pull
urlchecker::url_check()
devtools::build_readme()
devtools::check(remote = TRUE, manual = TRUE)
devtools::check_win_devel()
revdepcheck::revdep_check(num_workers = 4)
cran-comments.md
git push
Submit to CRAN:
usethis::use_version('major')
devtools::submit_cran()
Wait for CRAN...
usethis::use_github_release()
usethis::use_dev_version(push = TRUE)
In the documentation (intro vignette?) show a table with all plot and tab functions in rows, all parameters in cols and x when a function supports the parameter.
We still need three basic functions to fill the gap of the scheme:
And two plot functions for correlations
Add a markdown template with sensible defaults for creating reports:
RChecks: no visible binding for global variable
The .data pronoun does not work in select or rename...tidyselect does not like it.
In reports, respect the title in plot height calculation
codes <- codebook(ds)
write_xlsx(codes,"codebook.xlsx")
Dann manuell anpassen und:
codes <- read_xlsx("codebook-new.xlsx")
ds<- labs_apply(codebook)
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.