rempsyc - rempsyc 0.1.9

• New CRAN submission

rempsyc 0.1.8.3

• cormatrix_excel now relies entirely on correlation::cormatrix_to_excel() to reduce maintenance.

rempsyc 0.1.8.2

• best_duplicate: fix a bug with grouped tibbles leading to an error

rempsyc 0.1.8.1

• nice_density: add vjust = "inward" and hjust = "inward" to annotation to fix changes in ggpubr

Scientific Software - Peer-reviewed - R
Published by rempsyc over 1 year ago

rempsyc - rempsyc 0.1.8

• ✨Incoming✨

rempsyc 0.1.7.8

• nice_t_test: add temporary workaround for paired t-tests failing (#28)

rempsyc 0.1.7.7

• nice_table() gains a new argument: spacing (defaults to double space = 2, but can be set to 1 for single spacing)

rempsyc 0.1.7.6

• nice_lm() fix a bug with nice_lm() which doesn't work with lm models containing a single term (#27)

rempsyc 0.1.7.5

• nice_table() loses its nice_table class because of a name collision with the printing method of the afex package for nice_table objects (produced with afex::nice, #26)

rempsyc 0.1.7.4

• plot_means_over_time():
  • Now supports more than 5 groups (but shapes are not consistent after that)
  • Gains the groups.order argument, like nice_scatter(), to organize group factor levels in increasing or decreasing order.

rempsyc 0.1.7.3

• plot_means_over_time():
  • error bar colours changed from black to group colour to ease the visual distinction of a manual addition of vertical comparison significance bars.
  • New arguments to include significance bars and stars: significance_bars_x, significance_stars, significance_stars_x, significance_stars_y

rempsyc 0.1.7.2

• New function: grouped_bar_chart(), to plot categorical data options over several groups.

rempsyc 0.1.7.1

• New function: plot_means_over_time(), to plot group (intervention) data over time (1, 2, 3).

Scientific Software - Peer-reviewed - R
Published by rempsyc almost 2 years ago

rempsyc - rempsyc 0.1.7

• CRAN resubmission

rempsyc 0.1.6.2

• nice_table:
  • now converts sigma to its unicode counterpart ("\u03C3")
  • "z" and "Z" are now formatted to two digits instead of zero digits to comply with the Wald Z statistics in SEM (with lavaan and lavaanExtra)
  • now supports conversion of "CIlowerB" and "CIupperB" columns to "95% CI (b*)".
  • nice_t_test: argument warning becomes verbose to align with other packages.

rempsyc 0.1.6.1

• Breaking news: APA 7th edition actually advises against using beta to represent standardized coefficients, but suggests instead to use a lowercase italic b followed by an asterisk: b*. rempsyc will follow this APA recommendation going forward.

Scientific Software - Peer-reviewed - R
Published by rempsyc over 2 years ago

rempsyc - rempsyc 0.1.6

• CRAN submission

rempsyc 0.1.5.2

• nice_varplot: removed redundant legend

rempsyc 0.1.5.1

• nice_slopes and nice_mod:
  • Corrected bug when the data argument contained character variables when checking whether variables were standardized.

Scientific Software - Peer-reviewed - R
Published by rempsyc over 2 years ago

rempsyc - rempsyc 0.1.5

• CRAN submission

rempsyc 0.1.4.1

• nice_table:
  • Now automatically attempts to convert objects (e.g., emmGrid) to a dataframe so save users some headache.
  • Now automatically converts column name "t.ratio" to "t".
• get_dep_version: various improvements, now more robust

Scientific Software - Peer-reviewed - R
Published by rempsyc almost 3 years ago

rempsyc - rempsyc 0.1.4

• CRAN submission

rempsyc 0.1.3.5

• nice_table: fix bug when providing a table of more than one report::report object made from Wilcoxon tests.

rempsyc 0.1.3.4

• nice_lm_contrasts and nice_contrasts: Code simplified and now supports more than 3 groups, through easystats' modelbased package.

rempsyc 0.1.3.3

• nice_assumptions: p values from shapiro.test() are defined only for vectors of length between 3 and 5000 and was thus throwing an error in these cases. nice_assumptions now checks for the sample size and returns NA for samples outside this range (#17, @sjewo).

rempsyc 0.1.3.2

• New functions:
  • install_if_not_installed: will only install specified packages if they are not installed.
  • get_dep_version: checks the required version of a specific package dependency by looking at the specified package's DESCRIPTION file directly.
• best_duplicate: now supports the keep.rows argument, to add a column containing the original rows.
• overlap_circle: now support two other scoring methods: "percentage" (1 to 100), and "direct" (must provide the three values from circle 1, overlapping area, and circle 2).

rempsyc 0.1.3.1

• Added tests for all functions
• Accordingly, updated code to stop using deprecated functions from other packages (dplyr and ggplot2).

Scientific Software - Peer-reviewed - R
Published by rempsyc almost 3 years ago

rempsyc - rempsyc 0.1.3

rempsyc 0.1.3

• CRAN resubmission

rempsyc 0.1.2.5

• Fixed bug with check_installed where required package version were not being specified correctly.
• A way to install all dependencies at once are now provided in the README file.
• Contribution Guidelines are now available.

rempsyc 0.1.2.4

• nice_scatter: groups.order now also allows automatic grouping based on value for the response variable (one of the desired levels, "increasing", "decreasing", or "string.length"), for consistency with nice_violin.
• nice_violin: order.groups argument renames to groups.order for consistency with nice_scatter.

rempsyc 0.1.2.3

• nice_scatter:
  • fix bug whereas the group legend was lost even when using has.legend = TRUE.
  • has.legend now defaults to TRUE when using the group argument.
  • New argument method, with choices "lm" (default) or "loess".

rempsyc 0.1.2.2

• nice_violin:
  • For argument obs, there are two new choices: "dotplot" (same as obs = TRUE for backward compatibility) and "jitter", useful when there are a lot of observations.
  • Add new argument order.groups, to order the group factor levels based on value of the "response" variable, or on the string lengths for the group names (factor levels). Defaults to "none" to avoid changing previous behaviour.
  • Add new argument xlabels.angle, to tilt the labels of the x-axis (useful when working with long string names).

rempsyc 0.1.2.1

• nice_table: Correct beta rounding.

Scientific Software - Peer-reviewed - R
Published by rempsyc almost 3 years ago

rempsyc - rempsyc 0.1.2

• CRAN resubmission
• Readd leading zero for beta

rempsyc 0.1.1.9

• nice_table:
  • Now capitalize fit indices for a better integration with package {lavaanExtra}.
  • Simplification of internal coding and column formatting, e.g., columns names close equivalents, such as "pvalue" and "p.value", are now automatically converted to "p" (similar for others such as "chisq" to "chi2", conf.low and CI_low to CI_lower, etc.), for proper formatting.

rempsyc 0.1.1.8

• nice_table: It was requested that the leading zero be omitted for beta values (since it rarely goes beyond 1), so that is now the case (although note that in some edge cases it can go beyond 1, but that may indicate other problems with the model, such as high multicollinearity).

rempsyc 0.1.1.7

• format_r: will now convert NA values to "" instead of "NA"; this is useful when using the col.format.r argument in nice_table for correlation matrices.
• nice_table: CFI, TLI, RMSEA, and SRMR now omit the leading zero as they usually cannot be greater than one.

rempsyc 0.1.1.6

• Fix bug when checking column names (all functions).
• Fix bug in nice_violin to accept NULL group argument again (similar to problem corrected in version 0.1.1.4). Also changed the order of argument, with response second, and group third, since group becomes optional (as consistent with other similar functions).
• Fix a bug in citation("rempsyc").

rempsyc 0.1.1.5

• nice_lm and nice_lm_slopes: better error messages if the wrong object type is provided.

rempsyc 0.1.1.4

• Fix a newly introduced bug whereas nice_normality, nice_density, and nice_qq would not work without a group argument.

rempsyc 0.1.1.3

• nice_table and format_p: Now support the logical stars argument to add significance stars (defaults to TRUE for nice_table but FALSE for format_p).

rempsyc 0.1.1.2

• nice_table: It was pointed out that, in order to align with APA rules regarding tables, the first column should not be centered, but left-aligned. This has been corrected.
• nice_lm and nice_lm_slopes: Now attempt to automatically detect whether the variables were standardized, and if so, sets b.label = "B" automatically.
• nice_lm, nice_lm_slopes, nice_mod and nice_slopes: argument b.label becomes deprecated in favour of the new standardize argument, which defaults to FALSE for the first two (as this is most likely the desired result) and TRUE for the last two (because models are probably already specified as desired, so we do not want to refit them unless required).

rempsyc 0.1.1.1

• Now produces clearer error messages when specifying columns that don't exist, across all functions.
• sr2 function is now removed.

Scientific Software - Peer-reviewed - R
Published by rempsyc about 3 years ago

rempsyc - rempsyc 0.1.1

• CRAN resubmission!

rempsyc 0.1.0.9

• nice_assumptions: now supports list objects, and does not print the interpretation message anymore since it is available from the documentation.
• nice_var: now supports list objects.

rempsyc 0.1.0.8

• nice_violin and nice_contrasts: Now correctly handle missing values.
• New function! nice_lm_contrasts, to handle more complex contrast models involving e.g., interactions between variables (similar to nice_contrasts).

rempsyc 0.1.0.7

• nice_table:
  • footnote argument is renamed to simply note since this is what it is called in APA table language.
  • Fixed a bug whereas the note/footnote added an extra empty invisible row at the end.
  • Reduced cyclomatic complexity internally (swapped if statements for functions) dramatically (from 85 to 2).
• nice_contrasts
  • Default effect size changed to regular Cohen's d, and the robust Cohen's d (and other effects) can now be specified with the effect.type argument.

rempsyc 0.1.0.6

• Following easystats's policy of minimal reliance on external dependencies, we attempt to once again change a few more packages from hard to soft dependencies: flextable, effectsize, performance, insight, and methods. So we are left only with the basic blocks for coding: dplyr and rlang. For flextable, this is a breaking change for saving tables, since we now need to specify flextable::save_as_docx. However, this change makes sense because (1) we used flextable for only one function, (2) some users rely on rempsyc for plots and don't use tables, (3) this is consistent with how we are saving ggplot2 plots already, and (4) it also gives credit to the flextable package, as this is the powerhouse that produces the tables under the hood.
• nice_violin now provides more informative error messages if the response or group variables are misspelled.

rempsyc 0.1.0.5

• nice_lm, nice_mod, nice_lm_slopes, and nice_slopes now use "two.sided" as alternative for the sr2 effect size confidence interval, to facilitate interpretation and in accordance with current norms in psychology.

rempsyc 0.1.0.4

• Since easystats added a new function to calculate the sr2, effectsize::r2_semipartial, rempsyc now relies on that function. First, it has the advantage of also providing a confidence interval for the sr2. Second, it also fixes a bug when using factors in lm models that was introduced when switching from car::Anova to manual calculation via the formula interface.
• Accordingly, rempsyc::sr2 becomes deprecated and will be removed completely in the next major version, please use effectsize::r2_semipartial.
• nice_mod and nice_lm_slopes now use nice_lm internally to reduce code redundancy and shorten the code base (and nice_slopes uses nice_lm_slopes internally).
• To align with other package dependencies, rempsyc now requires R >= 3.6
• nice_reverse loses its warning parameter, as the warning seems unnecessary (and annoying to some), given all relevant information is available from the documentation.

rempsyc 0.1.0.3

• nice_table now only keeps the first occurrence of repeated (duplicated) dependent variables, and merges (and centers). Only works for a column called "Dependent Variables". For columns named "Predictor" and "Term", now also converts colons ":" to the multiplication symbol, "×".
• outliers_plot changes name to plot_outliers, to be consistent with the verb naming approach.

rempsyc 0.1.0.2

• nice_varplot and nice_var: Now ignores missing values when calculating variance.

rempsyc 0.1.0.1

• New function to visualize outliers: outliers_plot, which generates a violin scatter plot (dot plot) and adds lines for +/- 3 MAD (or SD, based on selected method). Importantly, supports plotting by group.
• nice_violin: at some point in time, it supported a single group (or no group), but this feature was lost at some point. It is now back.

Scientific Software - Peer-reviewed - R
Published by rempsyc over 3 years ago

rempsyc - rempsyc 0.1.0

• CRAN resubmission!

rempsyc 0.0.9.9

• nice_contrasts: corrects a bug whereas the effect sizes appeared in the wrong order.

rempsyc 0.0.9.8

• nice_violin gets some love after receiving its first citation (Jensen & Westergaard, 2022; https://doi.org/10.1111/lang.12525; yeah!!). It now defaults to not using bootstrapping per default, so it now needs to be requested explicitly. The bootstrapping method (BCa) is now specified in the documentation, and it is also clarified that the bootstrapping only applies to the confidence interval (not e.g., the mean). Underlying code is also simplified. Finally, a bug is fixed whereas the function used bootstrapping even when boot was set to FALSE.

rempsyc 0.0.9.7

• We move ggplot2 from a hard (imported) dependency to a soft (suggested) one, since many people seem to be using nice_table over the plotting features. And given the recent isoband CRAN dependency events.

rempsyc 0.0.9.6

• We get rid of car_Anova and lmsupport_modeleffectsizes and now compute the sr2 ourselves! With the help of our friends insight and performance.

rempsyc 0.0.9.5

• rempsyc gets lighter, as we get rid of the rather large car package dependency (by incorporating the single function we were using, Anova), and move other packages (boot, lmtest, ggrepel, ggsignif, and qqplotr) from required to suggested packages.

rempsyc 0.0.9.4

• nice_table: automatic formatting (of p-values, confidence intervals, etc.) now supports more than two levels of headers (with the separate.header argument).

rempsyc 0.0.9.3

• Now that openxlsx2 is on CRAN, cormatrix_excel2 replaces cormatrix_excel (and rempsyc package does not rely on an external GitHub dependency anymore, yeah!).

rempsyc 0.0.9.2

• nice_table: Multilevel headings, with the separate.header argument, now supports automatic formatting (of p-values, confidence intervals, etc.).

rempsyc 0.0.9.1

• nice_table:
  • fixed sr2 incorrectly showing leading zero.
  • new argument: col.format.ci, to format confidence intervals (accepts lists of lower and upper thresholds)
• find_mad and nice_assumptions: fixed printing method

Scientific Software - Peer-reviewed - R
Published by rempsyc over 3 years ago

rempsyc - rempsyc 0.0.9

• Initial CRAN submission!

Scientific Software - Peer-reviewed - R
Published by rempsyc over 3 years ago