I have not been able to reduce this to a minimal reproducible case, but when I'm working with a big RMarkdown document and I assign values to chunk options that are computed from variables defined/assigned in earlier chunks, I get problems when I try to run chunks interactively.
When I make small toy RMarkdown documents, I don't have this problem and having chunk options refer to variables defined in earlier chunks does not create these errors, so I am having trouble figuring out how to go from my big RMarkdown files to a small reproducible example.
Here is an example, excerpted from a larger document:
```{r block_variogram_plot, include=TRUE, dependson=c('make_spatial','load_data','interpolate')}
month_list <- c('JAN','FEB','MAR','APR','MAY','JUN','JUL','AUG','SEP','OCT','NOV','DEC')
var_labels <- c(precip = "Precipitation", t.max = expression(T[max]),
t.min = expression(T[min]), t.avg = expression(T[avg]))
cap_labels <- c(precip = "Precipitation", t.max = "$T_{\text{max}}$",
t.min = "$T_{\text{min}}$", t.avg = "$T_{\text{min}}$")
block_variogram_plot <- function(data, variable) {
variograms <- lapply(0:11, function(m) {
model_variogram(data[data$month == m,], formula(paste0(variable, " ~ 1")),
filter=TRUE)
})
v <- lapply(0:11, function(x) {
y <- variograms[[x+1]]$variogram; y$month = x; y
}) %>% bind_rows()
vx <- lapply(0:11, function(m) {
x <- variogramLine(variograms[[m+1]]$model,
maxdist = max(v$dist[v$month == m]))
x$month <- m; x
}) %>% bind_rows()
v$month <- factor(month_list[v$month + 1], levels = month_list, ordered = TRUE)
vx$month <- factor(month_list[vx$month + 1], levels = month_list, ordered = TRUE)
p <- ggplot(v, aes(x = dist, y = gamma)) + geom_point(size = I(3))
p <- p + geom_line(data = vx)
p <- p + facet_wrap( ~ month, scales = 'fixed')
invisible(p)
}
```{r plot_variograms, include=TRUE, fig.cap="Semivariograms with fitted linear models", fig.subcap=unlist(lapply(cap_labels, function(x) paste0("Semivariogram of ", x, " with linear model."))), dependson=c('block_variogram_plot')}
for (var in c('precip','t.max','t.min','t.avg')) {
var <- paste0(var, "_mean")
p <- block_variogram_plot(baseline, var)
plot(p + labs(x = "Distance (deg. lat/lon)", y = "Semivariance", title = var_labels[var]) +
theme_classic(base_size = 10))
# cat('\n\n')
}
If go to the plot_variograms chunk and interactively run all previous chunks (Ctrl+Alt+P), and then run the plot_variograms chunk with Ctrl+Shift+Enter or try to run the for loop by putting the cursor on the for line and doing Ctrl+Enter, I get an error message on the console I get the error message
Error in lapply(cap_labels, function(x) paste0("Semivariogram of ", x, :
object 'cap_labels' not found
However, cap_labels is defined in the interactive session:
> cap_labels
precip t.max t.min t.avg
"Precipitation" "$T_{\text{max}}$" "$T_{\text{min}}$" "$T_{\text{min}}$"
Further, RStudio sporadically crashes when I am doing this interactively, giving me a "Fatal error: Unexpected exception: bad allocation" message. The crashes are reasonably frequent but don't happen predictably enough for me to give a recipe for reproducing them.
I never had any of this trouble with RStudio 1.1 and earlier, so this is a new thing with the 1.2 previews.
Here is my SessionInfo:
> sessionInfo()
R version 3.5.1 (2018-07-02)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 17134)
Matrix products: default
locale:
[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252
attached base packages:
[1] grid stats graphics grDevices utils datasets methods base
other attached packages:
[1] bindrcpp_0.2.2 ggthemes_4.0.1 ggplot2_3.0.0 gstat_1.1-6 sp_1.3-1 readr_1.1.1
[7] purrr_0.2.5 stringr_1.3.1 dplyr_0.7.6 tidyr_0.8.1 readxl_1.1.0 rvest_0.3.2
[13] xml2_1.2.0 knitr_1.20
loaded via a namespace (and not attached):
[1] Rcpp_0.12.19 plyr_1.8.4 cellranger_1.1.0 pillar_1.3.0 compiler_3.5.1
[6] bindr_0.1.1 tools_3.5.1 xts_0.11-1 packrat_0.4.9-3 gtable_0.2.0
[11] tibble_1.4.2 lattice_0.20-35 pkgconfig_2.0.2 rlang_0.2.2 rstudioapi_0.8
[16] rgdal_1.3-4 withr_2.1.2 httr_1.3.1 hms_0.4.2 tidyselect_0.2.5
[21] spacetime_1.2-2 glue_1.3.0 R6_2.3.0 selectr_0.4-1 magrittr_1.5
[26] scales_1.0.0 intervals_0.15.1 assertthat_0.2.0 colorspace_1.3-2 stringi_1.2.4
[31] lazyeval_0.2.1 munsell_0.5.0 crayon_1.3.4 FNN_1.1.2.1 zoo_1.8-4
I have tried to paste the diagnostics report into this message, but I get an error about the body of my message being too long (the diagnostics report is 1.9 million characters) and I don't seem to be able to upload the diagnostics report because ".txt" is not a permitted file extension.