Tags: James-Thorson-NOAA/FishStatsUtils
Tags
Dev (#91) * change paste0(.) to file.path(.) in file naming for plots * another fix for `plot_factors` * fix bug in switchover from paste0 to file.path for plots * fix another bug in switchover from paste0 to file.path * fix marginaleffects * adding sample-based calculation of approx. sample-size * small expansion of docs * update paste0 to file.path * change default so fit_model runs in tempdir() ... ... to decrease change of user-error in re-using existing extrapolation-grid or Kmeans objects * separate plot_legend from plot_variable * adding new summary.fit_model option ... ... to output dataframe of expanded values * fix bug in amend_output when plotting EOF factors * small updates * switch from INLA to fmesher dependency * eliminate rgdal * bug fix for rotate_factors given sum-to-zero constraint * fix bug in updates for fmesher package * fix error from rnaturalearth using "sf" output * remove warning from inla.barrier.fem.copy * Update DESCRIPTION * run `document` --------- Co-authored-by: Jim Thorson <[email protected]>
Dev (#89) * small fix for problem flagged by Arnaud * adding extension step to get_latest_version .. ... to address @aallyn * another attempted fix * Fix buggy error-check for OSA residuals h/t Nicholas Ducharme-Barthe and #86 * fix bug in summary.fit_model * extend sample_variable to work with obj$env$parList(.) stuff * adding marginaleffects for visualizing covariate responses * fix bug in last commit * continue fiddling with marginaleffects * adding marginaleffects internal functions as S3methods * adding marginaleffects as DEPENDS to allow S3methods import * fix marginaleffects to work with map interface * trying to add map functionality to marginaleffects * fix marginaleffects to allow mapped parameters * fix bug when a_i uses areal units other than km2
Merging 2.12.0 * adding `plot_clusters` * removing deprecated functions from NAMESPACE * small fixes * fix bug in `plot_similarity` ... ... where it would throw an error when using a univariate model * improve plot_similarity ... ... to use smarter color-limits than the current default in corrplot package * update docs * fix plotting bugs in `plot_clusters` and `plot_similarity` * update docs for `plot_maps` * fix bug in last update to `plot_range_index` * fix `plot_clusters` to work well with NAs * add informative error for `plot_range_edge` ... ... for cases when normal approximation used by `sample_variable` performs poorly * prune dependencies using `renv::dependencies` * Try adding INLA to IMPORTS using `Additional_repositories` * moving SpatialDeltaGLMM fits to FishStatsUtils ... ... so that I can then remove SpatialDeltaGLMM from dependencies of VAST * noting `summarize_covariance` is deprecated * housekeeping: remove .onAttach action, and deprecate map_hypervariance * eliminate dependency `gap` * small fixes * adding Lindegren et al. ref to `plot_clusters` * fix obj$report() when using intern=TRUE * improved tracking for run times * update messages * fix bug * Deprecate Calc_Anisotropic_Mesh * a * fix bug in plot_similarity ... ... where plot.hclust(.) throws an error with only 2 groups * Update version number * adding `form` to plot.fit_model ... .. so that residuals can be easily plotted by category * improve `plot_clusters` and `summary.fit_model` ... ... in support of h/t Joe Bizzarro and West Coast diet analysis * update docs * fix `plot_similarity` bug ... ... which threw uniformative error when doing EpsilonTime for a model with only one time-interval * better fix for plot_similarity .. ... when n_t = 1 * add warning for incorrect class in `plot_variable` * adding `plot_residual_semivariance` ... ... as additional diagnostic for model fit * small updates * fix typo in `plot_residual_semivariance` * Update fit_model ... ... to allow user to pass data_list, spatial_list, or extrapolation_list and thereby override default functions. See #83 * fix bug in `plot_residual_semivariance` and ... ... add backdoor for new maps for barrier feature, see #84 * fix bug in last commit * Improve `plot_results` ... ... by allowing `Expansion_cz` to be passed to `calculation_proportion` * adding sf option for plotting maps * adding new epsilon-correction by default * small plotting fix for format="sf" ... ... h/t Sarah Gaichas for flagging the issue * adding demo code for SVC paper * fix to work with suplus-production example to use Bratio and Fratio * increase eval.max and iter.max defaults * various doc updates
Dev (#77) * Improve rotate_factors... ... to allow working with projected values of factors * small fix to plot_factors update... ... h/t Kristen Omori * fix bug with many extrapolation-grid cells with Area=0... ... this issue arises e.g. with Region = "Other" and h/t Ellen Yasumiishi for flagging the issue! the bug manifested in an uniformative error, so no change of previous users having untrustworthy results arising from it. * fix bug in last commit * update docs * improve docs * Enhance rotate_factors ... ... to work even when estimating a factor model where some factors have variance approaching zero. * fix a plotting bug in `plot_factors` ... ... that arose when plotting factor-model Omegas * fix to `plot_factors` when doing dual ordination * Fix summary.fit_model( ..., what="residuals") ... ... in response to #72, h/t Nicholas Ducharme-Barthe for reporting the bug (which threw an uninformative error) * Adding plots for spatially-varying catchability * improve panel-labels in density plots * add loc_s and latlon_s ... ... for potential future use in new feature for habitat-specific variance-covariance among variables * fix typo in last commit * fix typo in last commit * Change default colors for plots ... ... and add option to calculate subarea totals using Longitude for Region="eastern_bering_sea" * adding units package * update docs * adding amend_output ... ... which labels Report for use in subsequent plotting * Fix bug in fit_model ... ... caused during earlier push * Improve `amend_output` * Update plot_biomass_index ... ... to use metadata_ctz, and deprecate compatibility with MIST and SpatialDeltaGLMM * fix new bug in plot_biomass_index ... ... introduced a couple commits earlier in dev branch; also improve stability for DHARMa using PIT residuals * Continue to improve stability for DHARMa residuals ... ... see discussion here: florianhartig/DHARMa#286 * update plots * Change make_covariates ... ... to avoid passing unnecessarily large X_itp object, which exceeded R memory limits in some extreme cases * small update ... ... to deal with amend_output change where Density=0 such that log(Density) = -Inf, which was causing plotting problems * Update deprecated.R functions for old diagnostic plots ... ... to deal with case of not specifying explicit units * various updates to doxygen reference docs * Update deprecated.R * Update deprecated.R * fix DHARMa residuals when b_i=NA, and improve plot_data(.) to work with projargs * fix bug in dev-changes for project_coordinates * fix bugs in dev-branch for plot_data * improve DHARMa plots ... ... by removing useless p-values * fix dev-branch bug in plot_biomass_index ... ... which now requires `extrapolation_list` to pull appropriate units for plot * adding West Coast Groundfish combo survey ... ... to VAST shapefiles * small style-guide improvements * Improve default labelling in plots ... ... by calling `amend_output` within `plot_maps` * Avoid crash in `plot_maps` ... ... when asking to plot variable that is excluded from model * Fix annoying warnings about PROJ.4 * Eliminate (most) annoying warnings from PROJ.4 * small fixes .. ... to eliminate remaining PROJ.4 warnings, and fix plotting crash regarding contours in missing variables * small fix ... ... to allow for passing a color function `col` to `plot.fit_model` * remove dependency `plotKML` with `raster` h/t @Jason-Conner-NOAA for pointing out that `plotKML` was removed from CRAN * fix bug in last commit * Simplify plot_biomass_index ... ... by incorporating `amend_output` into function * fit issues in last committ ... ... such that amend_output is now responsible for handing units * adding units for mean_Z_ctm ... ... extracted using `sf` which is now a dependency * Fix use of units for range-shift metrics * fix bug in recent commit ... ... which caused `plot_maps` to not plot SEs properly by default * update docs * fix bug in amend_outputs ... ... regarding optional Report slot `mean_D_ctl` * update docs
Dev (#77) * Improve rotate_factors... ... to allow working with projected values of factors * small fix to plot_factors update... ... h/t Kristen Omori * fix bug with many extrapolation-grid cells with Area=0... ... this issue arises e.g. with Region = "Other" and h/t Ellen Yasumiishi for flagging the issue! the bug manifested in an uniformative error, so no change of previous users having untrustworthy results arising from it. * fix bug in last commit * update docs * improve docs * Enhance rotate_factors ... ... to work even when estimating a factor model where some factors have variance approaching zero. * fix a plotting bug in `plot_factors` ... ... that arose when plotting factor-model Omegas * fix to `plot_factors` when doing dual ordination * Fix summary.fit_model( ..., what="residuals") ... ... in response to #72, h/t Nicholas Ducharme-Barthe for reporting the bug (which threw an uninformative error) * Adding plots for spatially-varying catchability * improve panel-labels in density plots * add loc_s and latlon_s ... ... for potential future use in new feature for habitat-specific variance-covariance among variables * fix typo in last commit * fix typo in last commit * Change default colors for plots ... ... and add option to calculate subarea totals using Longitude for Region="eastern_bering_sea" * adding units package * update docs * adding amend_output ... ... which labels Report for use in subsequent plotting * Fix bug in fit_model ... ... caused during earlier push * Improve `amend_output` * Update plot_biomass_index ... ... to use metadata_ctz, and deprecate compatibility with MIST and SpatialDeltaGLMM * fix new bug in plot_biomass_index ... ... introduced a couple commits earlier in dev branch; also improve stability for DHARMa using PIT residuals * Continue to improve stability for DHARMa residuals ... ... see discussion here: florianhartig/DHARMa#286 * update plots * Change make_covariates ... ... to avoid passing unnecessarily large X_itp object, which exceeded R memory limits in some extreme cases * small update ... ... to deal with amend_output change where Density=0 such that log(Density) = -Inf, which was causing plotting problems * Update deprecated.R functions for old diagnostic plots ... ... to deal with case of not specifying explicit units * various updates to doxygen reference docs * Update deprecated.R * Update deprecated.R * fix DHARMa residuals when b_i=NA, and improve plot_data(.) to work with projargs * fix bug in dev-changes for project_coordinates * fix bugs in dev-branch for plot_data * improve DHARMa plots ... ... by removing useless p-values * fix dev-branch bug in plot_biomass_index ... ... which now requires `extrapolation_list` to pull appropriate units for plot * adding West Coast Groundfish combo survey ... ... to VAST shapefiles * small style-guide improvements * Improve default labelling in plots ... ... by calling `amend_output` within `plot_maps` * Avoid crash in `plot_maps` ... ... when asking to plot variable that is excluded from model * Fix annoying warnings about PROJ.4 * Eliminate (most) annoying warnings from PROJ.4 * small fixes .. ... to eliminate remaining PROJ.4 warnings, and fix plotting crash regarding contours in missing variables * small fix ... ... to allow for passing a color function `col` to `plot.fit_model` * remove dependency `plotKML` with `raster` h/t @Jason-Conner-NOAA for pointing out that `plotKML` was removed from CRAN * fix bug in last commit * Simplify plot_biomass_index ... ... by incorporating `amend_output` into function * fit issues in last committ ... ... such that amend_output is now responsible for handing units * adding units for mean_Z_ctm ... ... extracted using `sf` which is now a dependency * Fix use of units for range-shift metrics * fix bug in recent commit ... ... which caused `plot_maps` to not plot SEs properly by default * update docs * fix bug in amend_outputs ... ... regarding optional Report slot `mean_D_ctl` * update docs
PreviousNext