ampiccinin
commented
7 years ago @andkov : NuAge, LASA and ILSE are also missing from Gait, but please fix Grip first, then check through all the links for the more recently created pages.
Closed ampiccinin closed 7 years ago
ampiccinin
commented
7 years ago @andkov : NuAge, LASA and ILSE are also missing from Gait, but please fix Grip first, then check through all the links for the more recently created pages.
andkov
commented
7 years ago Thanks for the detail instructions, @ampiccinin. I'll get on it today.
andkov
commented
7 years ago @ampiccinin, @andreazammit, just to keep you posted, I"m looking at the code. The reports fail to reproduce. Which is not entirely unexpected, but certaintly frustraiting. No cause for worries, just requires a bit more unanticipated (when it really is?) code maintanance on my side. I couldn't do a quick fix I was hoping for: it wasn't just a link issue, the reprot produced a wrong table and i missed it. I think because we were focusing on pulmonary then and grip was sort of a side effect. So i'll have to dive deeply in order to address it, which is a good thing because it's about time I do. I'll take care of the issues listed first, though.
@ampiccinin , not all studies contributed to all measures. I was adjusting the table of links. (just to clarify, we are talking about this page: https://github.com/IALSA/IALSA-2015-Portland/tree/master/projects/grip-cognitive). So I interpret the missingness of ILSE, LASA and NuAge as meaningful: the models contributed to the pool did not usethis predictor. I"ll be glad to double check; this will also be a good opportunity to remind us what studies contribute to what track.
domain map is not working because the graph has not been produced yet. This is on my agenda for the Grand Cleaning of IALSA repositories i'm scheduling for this next week.
RADC/MAP - yes, thanks for picking up on these things. Some of these works are leftovers from the earlier times, when I did not know what I was doing to a greater extent. I think i was thinking "RADC" was more general, but I see now why MAP is a better choice. I"ll keep an eye on this as I clean.
andkov
commented
7 years ago @ampiccinin , thanks for pointing out the incorrect study names on project READMEs. I think I was focusing on pulmonary at a time and didn't check the materials for other tracks to save time. I'm coming back to the front face of track READMEs to ensure the availability of all materials for publications.
To offer a bird-eye glimpse at the current state of the catalog (the data object holding all processed and groomed model results) at this time, here are the tallies of models from each study, broken down by track. I'll use these lists to design each track's readme.
> catalog %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("gait","tug")) %>%
+ print(n=nrow(.))
Source: local data frame [7 x 3]
Groups: process_a [2]
process_a study_name n_models
<chr> <chr> <int>
1 gait eas 120
2 gait elsa 6
3 gait hrs 18
4 gait lasa 6
5 gait map 152
6 gait octo 104
7 tug ilse 43
> catalog %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("grip")) %>%
+ print(n=nrow(.))
Source: local data frame [9 x 3]
Groups: process_a [1]
process_a study_name n_models
<chr> <chr> <int>
1 grip eas 131
2 grip elsa 4
3 grip hrs 18
4 grip ilse 10
5 grip lasa 6
6 grip map 152
7 grip nuage 20
8 grip octo 104
9 grip satsa 56
> catalog %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("fev","pef")) %>%
+ print(n=nrow(.))
Source: local data frame [8 x 3]
Groups: process_a [2]
process_a study_name n_models
<chr> <chr> <int>
1 fev elsa 4
2 fev map 152
3 fev nas 66
4 fev satsa 106
5 pef eas 49
6 pef hrs 18
7 pef lasa 6
8 pef octo 104b1 aehplus modelsAnd here are the same tables, but NOW counting only the bivariate linear models of type aehplus, the one that we use in reporting correlations.
> catalog %>%
+ filter(model_number=="b1",model_type =="aehplus") %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("gait","tug")) %>%
+ print(n=nrow(.))
Source: local data frame [7 x 3]
Groups: process_a [2]
process_a study_name n_models
<chr> <chr> <int>
1 gait eas 22
2 gait elsa 6
3 gait hrs 6
4 gait lasa 6
5 gait map 38
6 gait octo 26
7 tug ilse 11
> catalog %>%
+ filter(model_number=="b1",model_type =="aehplus") %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("grip")) %>%
+ print(n=nrow(.))
Source: local data frame [9 x 3]
Groups: process_a [1]
process_a study_name n_models
<chr> <chr> <int>
1 grip eas 22
2 grip elsa 4
3 grip hrs 6
4 grip ilse 10
5 grip lasa 6
6 grip map 38
7 grip nuage 4
8 grip octo 26
9 grip satsa 20
> catalog %>%
+ filter(model_number=="b1",model_type =="aehplus") %>%
+ group_by(process_a, study_name) %>%
+ summarize(n_models=n()) %>%
+ filter(process_a %in% c("fev","pef")) %>%
+ print(n=nrow(.))
Source: local data frame [8 x 3]
Groups: process_a [2]
process_a study_name n_models
<chr> <chr> <int>
1 fev elsa 4
2 fev map 38
3 fev nas 8
4 fev satsa 20
5 pef eas 22
6 pef hrs 6
7 pef lasa 6
8 pef octo 26@ampiccinin , @andreazammit all reports in GRIP has been updated and links corrected. You can access all of them at the project's README: https://github.com/IALSA/IALSA-2015-Portland/tree/master/projects/grip-cognitive
@andkov - it has just been noted that some pieces of the grip-cognitive area are being haunted by the pulmonary outputs as well as some other issues. Andrea Zammit is trying to access these now to complete the grip-cog paper, so the sooner you could correct these the better. Thanks!
*( also - please correct the study associated with Chenkai's name. It was initially HABC, but now it is HRS. To be thorough, it would be worth noting that initially Chenkai's plan was to use HABC, but the way we have it listed now looks like an error).
Similarly, the MAP study is still labeled "RACD" on our main Portland page. I tried to edit these, but apparently do not have permission (or don't understand the "edit" button...). On some of the sub-pages, RADC has been edited to MAP, but the alphabetical ordering has not been updated (sorry to be so picky, but it can be confusing).
Thanks!