search

dcomtois / summarytools

R Package to Quickly and Neatly Summarize Data
504 stars 77 forks source link

dfSummary() generates empty plot on Windows #94

Closed GegznaV closed 4 years ago

GegznaV commented 4 years ago

Either dfSummary() function or, most probably, its print() method generates an empty plot. The plot shouldn't be created at all.

image

If graph.col = FALSE is used, the plot is not generated. But the "plots" column is also missing.

This behavior (empty unnecessary plot) should be fixed.

GegznaV commented 4 years ago

Warning NAs introduced by coercion is also present. It should not be present.

dcomtois commented 4 years ago

For the empty plot, it will be fixed in upcoming version. For the NA's, could you provide a reproducible example pls? Thx.

On Mon., Oct. 7, 2019, 15:18 Vilmantas Gegzna, notifications@github.com wrote:

Warning NAs introduced by coercion is also present. It should not be present.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/dcomtois/summarytools/issues/94?email_source=notifications&email_token=ABNVLA62PI4HMW3PSUMEA33QNODRFA5CNFSM4I6IHDUKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEARPXJQ#issuecomment-539163558, or mute the thread https://github.com/notifications/unsubscribe-auth/ABNVLAZHOSZ754VJYX4YUFTQNODRFANCNFSM4I6IHDUA .

GegznaV commented 4 years ago

It seems that NA message appears only in R Markdown document.

``` {r setup}
library(tidyverse)
library(summarytools)
```

```{r}
dfSummary(iris, graph.col = FALSE)
```

image

dcomtois commented 4 years ago

Can you pls retry after installing the dev-current version from github?

devtools::install_github("dcomtois/summarytools", ref = "dev-current")

GegznaV commented 4 years ago 
devtools::install_github("dcomtois/summarytools", ref = "dev-current")

Skipping install of 'summarytools' from a github remote, the SHA1 (01394c8d) has not changed since last install. Use force = TRUE to force installation

Both empty plot and NA issues still persist.

dcomtois commented 4 years ago

Sorry about the delay... Could you try and check the value of the following option pls?

st_options('use.x11')

If it's FALSE, you could try and set it to TRUE and see what happens:

st_options(use.x11 = TRUE)

GegznaV commented 4 years ago

st_options('use.x11') was TRUE. If I disable it, no empty plot is present:

image

But I do not understand: isn't x11 a graphical device for Unix-like OS'es, especially Linux? Shouldn't it be disabled on Windows by default?

dcomtois commented 4 years ago

See this page for details about x11 on Windows: https://stat.ethz.ch/R-manual/R-devel/library/grDevices/html/x11.html

As for the issue still persisting, I'm not sure what to make of it... I'd try forcing the reinstallation if possible.

dcomtois commented 4 years ago

Closing as this is likely resolved, if not just let me know and I'll reopen.

gregkvas commented 4 years ago

I've got the same issues described by GegznaV when using summarytools in R notebooks (Windows 10 Pro Vers. 1803; RStudio Vers. 1.2.1335; R Vers. 3.5.3; summarytools Vers. 0.9.6)

After setting st_options(use.x11 = FALSE) the empty plot is NOT generated anymore, thx for that, but the output NAs introduced by coercion still remains (screenshot).

image