Error: cogdf must be a cognostics object - call as_cognostics() to cast it as such.

[stefvanbuuren]

The following code was previously - in May 2020 - working (see #25):

library(trelliscopejs)
library(ggplot2)
library(dplyr)
library(tidyr)

my_order <- c("subcompact", "compact", "midsize",
              "minivan", "pickup", "suv", "2seater")

# Ryan's solution: https://github.com/hafen/trelliscopejs/issues/25
mpg %>%
  mutate(class = factor(class, levels = my_order),
         class_index = as.integer(class)) %>%
  group_by(class, class_index) %>%
  nest() %>%
  mutate(panel = map_plot(data,
                          ~ ggplot(data = .x, aes(displ, hwy)) +
                            geom_point() +
                            xlab("Engine displacement, litres") +
                            ylab("Highway miles per gallon") +
                            xlim(1, 7) + ylim(10, 60) +
                            theme_light())) %>%
  trelliscope(name = "MPG custom panel order",
              state = list(sort = list(sort_spec("class_index"))),
              nrow = 1, ncol = 1,
              path = "docs/mpg-custom")

In R 4.0.3 and trelliscopejs 0.2.5 it throws

58% 7/12 eta: 1sError: cogdf must be a cognostics object - call as_cognostics() to cast it as such.
In addition: Warning message:
`filter_()` is deprecated as of dplyr 0.7.0.
Please use `filter()` instead.
See vignette('programming') for more help
This warning is displayed once every 8 hours.
Call `lifecycle::last_warnings()` to see where this warning was generated. 

The call writes contents into the appfiles directory, but produces no index.html files or lib directory.

Could this be related to #91? Any work-arounds?

[stefvanbuuren]

> sessionInfo()
R version 4.0.3 (2020-10-10)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 10.16

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

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

other attached packages:
[1] tidyr_1.1.2         dplyr_1.0.2         ggplot2_3.3.2       trelliscopejs_0.2.5

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.5              later_1.1.0.1           pillar_1.4.6            compiler_4.0.3          prettyunits_1.1.1       base64enc_0.1-3         tools_4.0.3            
 [8] progress_1.2.2          digest_0.6.27           mclust_5.4.6            jsonlite_1.7.1          lifecycle_0.2.0         tibble_3.0.4            gtable_0.3.0           
[15] checkmate_2.0.0         pkgconfig_2.0.3         rlang_0.4.9             shiny_1.5.0             rstudioapi_0.13         xfun_0.19               fastmap_1.0.1          
[22] withr_2.3.0             knitr_1.30              generics_0.1.0          vctrs_0.3.5             DistributionUtils_0.6-0 hms_0.5.3               grid_4.0.3             
[29] webshot_0.5.2           tidyselect_1.1.0        glue_1.4.2              R6_2.5.0                farver_2.0.3            purrr_0.3.4             magrittr_2.0.1         
[36] promises_1.1.1          scales_1.1.1            autocogs_0.1.3          backports_1.2.0         ellipsis_0.3.1          htmltools_0.5.0         xtable_1.8-4           
[43] mime_0.9                colorspace_2.0-0        httpuv_1.5.4            labeling_0.4.2          munsell_0.5.0           crayon_1.3.4.9000  

[stefvanbuuren]

 traceback()
6: stop(paste(strwrap(paste(...), exdent = 7), collapse = "\n"), 
       call. = FALSE)
5: stop_nice("cogdf must be a cognostics object -", "call as_cognostics() to cast it as such.")
4: write_display_obj(cog_df, panel_example = panels[[1]], panel_img_col = panel_img_col, 
       base_path = params$path, id = params$id, name = params$name, 
       group = params$group, desc = desc, height = height, width = width, 
       md_desc = md_desc, state = params$state, jsonp = params$jsonp, 
       split_sig = params$split_sig, self_contained = params$self_contained, 
       thumb = params$thumb, pb = pb)
3: trelliscope.data.frame(., name = "MPG custom panel order", state = list(sort = list(sort_spec("class_index"))), 
       nrow = 1, ncol = 1, path = "docs/mpg-custom")
2: trelliscope(., name = "MPG custom panel order", state = list(sort = list(sort_spec("class_index"))), 
       nrow = 1, ncol = 1, path = "docs/mpg-custom")
1: mpg %>% mutate(class = factor(class, levels = my_order), class_index = as.integer(class)) %>% 
       group_by(class, class_index) %>% nest() %>% mutate(panel = map_plot(data, 
       ~ggplot(data = .x, aes(displ, hwy)) + geom_point() + xlab("Engine displacement, litres") + 
           ylab("Highway miles per gallon") + xlim(1, 7) + ylim(10, 
           60) + theme_light())) %>% trelliscope(name = "MPG custom panel order", 
       state = list(sort = list(sort_spec("class_index"))), nrow = 1, 
       ncol = 1, path = "docs/mpg-custom")

[hafen]

Can you try this and see if it works:

mpg %>%
  mutate(class = factor(class, levels = my_order),
         class_index = as.integer(class)) %>%
  group_by(class, class_index) %>%
  nest() %>%
  mutate(panel = map_plot(data,
                          ~ ggplot(data = .x, aes(displ, hwy)) +
                            geom_point() +
                            xlab("Engine displacement, litres") +
                            ylab("Highway miles per gallon") +
                            xlim(1, 7) + ylim(10, 60) +
                            theme_light())) %>%
  ungroup() %>%
  trelliscope(name = "MPG custom panel order",
              state = list(sort = list(sort_spec("class_index"))),
              nrow = 1, ncol = 1,
              path = "docs/mpg-custom")

Note the only difference is the ungroup(). This should not be necessary so I'll introduce a fix for it. In the mean time, if the ungroup gives you the desired result, that will work for now.

[stefvanbuuren]

Ryan, that's great. Adding ungroup() %>% prior to trelliscope() solves the problem, both in the basic mpg example as well as in my production plots.

Thanks a lot.