Open peterhurford opened 8 years ago
If you (a) run the entire model and then (b) define a new stage and then (c) try to run that stage, you get this very unhelpful error:
> run("imsurvey", "analyze/involved_local_EA x donate") Loading model: models/dev/imsurvey Copying cached environments... Beginning 3. analyze stage... Running 10. involved_local_EA x donate... Error: is.environment(to) && is.environment(from) is not TRUE In environment imports:Ramd: run("imsurvey", "analyze/involved_local_EA x donate") In /Users/peterhurford/.R/.syberia/engines/github/robertzk/modeling.sy/master/config/global/modeling/run_model.R:27: invisible(director$cache_set("last_run", director$resource(keys[1])$run(..., verbose = verbose))[[1L]]) In package director: self$cache$set(k, v) In package base: structure(list(value), .Names = key) In package stagerunner: director$resource(keys[1])$run(..., verbose = verbose) In package stagerunner: run_stage(..., remember_flag = TRUE) In package stagerunner: stages[[stage_index]]$run(stage_key[[stage_index]], normalized = TRUE, verbose = verbose, .depth = .depth + 1, ...) In package stagerunner: self$.before_env(stage_index) In package objectdiff: package_function("objectdiff", "force_push")(self$.context, current_commit) In package objectdiff: replay(env, commit, silent = TRUE) In package objectdiff: copy_env(env %$% env, (env %$% reference)[[reference_index]]) In package base: stopifnot(is.environment(to) && is.environment(from)) In package base: stop(sprintf(ngettext(length(r), "%s is not TRUE", "%s are not all TRUE"), ch), call. = FALSE, domain = NA) Error: is.environment(to) && is.environment(from) is not TRUE
Mirroring this as a coalescing issue in stagerunner.
If you (a) run the entire model and then (b) define a new stage and then (c) try to run that stage, you get this very unhelpful error: