issues
search
garner-code
/
PSY2R
translating PSY to R
3
stars
16
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
In and out files for Psy with repeated measures ANOVA
#62
MashaPtukha
opened
1 week ago
0
Adding an R markdown file called PsyR_info
#61
MashaPtukha
closed
1 week ago
0
added a function to create contrasts for interaction
#60
Viiiccckkky
closed
1 week ago
0
added names to README
#59
Viiiccckkky
closed
1 week ago
0
Update README.md
#58
carmzh
closed
1 week ago
0
Addes an R script called types_of_anovas
#57
MashaPtukha
closed
1 week ago
0
Whether Bonferroni correction behaves the same for between, within, and between x within contrasts
#56
carmzh
closed
1 week ago
0
how does emmeans::recover_data(mod) behave when mod has been defined by different methods to perform ANOVA?
#55
garner-code
closed
5 days ago
0
function required to add CIs
#54
garner-code
closed
5 days ago
0
5. check how SPM defines contrasts
#53
carmzh
opened
1 week ago
0
4. warning message
#52
carmzh
closed
1 week ago
0
3. checking and extracting info from ANOVA model
#51
carmzh
closed
1 week ago
1
2. how emmeans behaves when there are multiple within factors
#50
MashaPtukha
opened
1 week ago
0
need some example code on how to test variables and correctly throw errors or warnings when tests fal
#49
garner-code
opened
1 week ago
1
need some example code on how to check if an object or variable exists
#48
garner-code
closed
1 week ago
1
scaling contrast vectors
#47
Viiiccckkky
opened
1 week ago
0
do all aov models have the same fields in the objects? e.g. Den DF?
#46
garner-code
closed
1 week ago
1
1. compare each to each and one to many
#45
MashaPtukha
opened
1 week ago
1
test emmeans behaviour with different ways of defining the ANOVA model (e.g. aov_ez, ezANOVA,aov)
#44
garner-code
closed
1 week ago
0
Work on 'getting_Psy_contrast_estimates_from_emmeans.R' to be a general wrapper for getting contrast tables, to add the Psy business to
#43
garner-code
closed
5 days ago
0
replicate workflows with user provided data (see other workflows comment for details)
#42
garner-code
opened
1 week ago
0
Have a workflow for using the existing functions
#41
garner-code
closed
1 week ago
1
first draft of generalsing compute_win_se
#40
jorittmo
closed
4 months ago
1
emmeans_info_folder
#39
krogge-obando
closed
4 months ago
0
first draft of the gcr_crit function which calculates gcr critical value
#38
jorittmo
closed
4 months ago
1
How do other R packages organise their functions?
#37
garner-code
closed
4 months ago
1
put functions together into a package
#36
garner-code
closed
1 week ago
0
If we have time, get into Boik's smr distribution
#35
garner-code
opened
5 months ago
0
Tinkering under the hoods: how does emmeans interact with model objects in R?
#34
garner-code
closed
1 week ago
0
Recycling is good! Work out which parts of Psy are already covered by emmeans and MBESS
#33
garner-code
closed
1 week ago
2
Work out best method to implement computation of critical GCR under the null hypothesis (the keystone & unicorn)
#32
garner-code
closed
1 week ago
4
A to-do list before I even get to Brainhack
#31
garner-code
closed
4 months ago
0
Finish to-do notes here on pascal to R implementation of GCR functions
#30
garner-code
closed
5 months ago
0
Submit project to Brainhack
#29
garner-code
closed
4 months ago
0
adding notes on bird 2005
#28
lydiabarnes01
closed
11 months ago
0
End of Hackathon Progress by Group to Convert Pascal to Python
#27
GalaxyBound
closed
11 months ago
0
Attempted int SEs
#26
katmarchenko
closed
11 months ago
0
t-Individual function and implementation added
#25
katmarchenko
closed
11 months ago
0
t_Individual
#24
katmarchenko
closed
11 months ago
0
adding bonferroni alternate example
#23
lydiabarnes01
closed
11 months ago
0
"updating bonferroni CI example file name"
#22
lydiabarnes01
closed
11 months ago
0
added t-Bonferroni function and implementation
#21
katmarchenko
closed
11 months ago
0
alternate example bonferroni CI implementation
#20
lydiabarnes01
closed
11 months ago
0
KM implementation edit
#19
katmarchenko
closed
11 months ago
0
why is the n df residual for between contrast sigma, but n subjects for within?
#18
garner-code
opened
11 months ago
0
why slight disparity for between subjects post-hoc contrasts?
#17
garner-code
closed
11 months ago
0
adding rough bonferroni ci method for between-subjects comparisons
#16
lydiabarnes01
closed
11 months ago
0
is there any need to correct for sphericity?
#15
garner-code
closed
5 months ago
0
updated code
#14
julieechow
closed
11 months ago
0
Rcode
#13
julieechow
closed
11 months ago
0
Next