IALSA / IALSA-2015-Portland

Hub to accompany IALSA 2015 workshop at Portland, OR, Feb 22-25, 2015
GNU General Public License v2.0
2 stars 0 forks source link

Basic graphs for Physical Track #96

Open andkov opened 8 years ago

andkov commented 8 years ago

@ampiccinin , @GracielaMuniz , @annierobi , the first graph based on the results.csv has been produced (Screen below). You can see all graphs for the available pairs in ./reports/physical/basic_graphs/figure_rmd or as a md or html document.

gait_grip-1

andkov commented 8 years ago

New report give the scatter plots of the factor scores extracted from .gh5 files. All graphs are labeled and stored in ./reports/physical/fscores_scatter/figure_rmd folder. You can view them as md or html stored on github.

radc_female_aehplus_gait_grip-1

andkov commented 8 years ago

@ampiccinin , the graphs for all available . gh5 files has been uploaded. See this README to view them online, however, I recommend syncing and viewing their local versions at ./reports/physical/fscores_scatter/figure_rmd

At this moment, the graphs depict the following models:

 [1] "./studies/eas/physical/b1_female_aehplus_grip_gait.gh5"  
 [2] "./studies/eas/physical/b1_female_aehplus_grip_pef.gh5"   
 [3] "./studies/eas/physical/b1_female_aehplus_pef_gait.gh5"   
 [4] "./studies/eas/physical/b1_male_aehplus_grip_gait.gh5"    
 [5] "./studies/eas/physical/b1_male_aehplus_grip_pef.gh5"     
 [6] "./studies/eas/physical/b1_male_aehplus_pef_gait.gh5"    

 [7] "./studies/hrs/physical/b1_female_aehplus_grip_gait.gh5"  
 [8] "./studies/hrs/physical/b1_female_aehplus_grip_pef.gh5"   
 [9] "./studies/hrs/physical/b1_female_aehplus_pef_gait.gh5"   
[10] "./studies/hrs/physical/b1_male_aehplus_grip_gait.gh5"    
[11] "./studies/hrs/physical/b1_male_aehplus_grip_pef.gh5"     
[12] "./studies/hrs/physical/b1_male_aehplus_pef_gait.gh5"     

[13] "./studies/lasa/physical/b1_female_aehplus_gait_grip.gh5" 
[14] "./studies/lasa/physical/b1_female_aehplus_pek_gait.gh5"  
[15] "./studies/lasa/physical/b1_female_aehplus_pek_grip.gh5"  
[16] "./studies/lasa/physical/b1_male_aehplus_gait_grip.gh5"   
[17] "./studies/lasa/physical/b1_male_aehplus_pek_gait.gh5"    
[18] "./studies/lasa/physical/b1_male_aehplus_pek_grip.gh5"    

[19] "./studies/octo/physical/b1_female_aehplus_gait_grip.gh5" 
[20] "./studies/octo/physical/b1_female_aehplus_pek_gait.gh5"  
[21] "./studies/octo/physical/b1_female_aehplus_pek_grip.gh5"  
[22] "./studies/octo/physical/b1_male_aehplus_gait_grip.gh5"   
[23] "./studies/octo/physical/b1_male_aehplus_pek_gait.gh5"    
[24] "./studies/octo/physical/b1_male_aehplus_pek_grip.gh5"   

[25] "./studies/radc/physical/b1_female_aehplus_fev_gait.gh5"  
[26] "./studies/radc/physical/b1_female_aehplus_fev_grip.gh5"  
[27] "./studies/radc/physical/b1_female_aehplus_gait_grip.gh5" 
[28] "./studies/radc/physical/b1_male_aehplus_fev_gait.gh5"    
[29] "./studies/radc/physical/b1_male_aehplus_fev_grip.gh5"    
[30] "./studies/radc/physical/b1_male_aehplus_gait_grip.gh5"   

[31] "./studies/satsa/physical/b1_female_aehplus_gait_fev.gh5" 
[32] "./studies/satsa/physical/b1_female_aehplus_gait_grip.gh5"
[33] "./studies/satsa/physical/b1_female_aehplus_grip_fev.gh5" 
[34] "./studies/satsa/physical/b1_male_aehplus_gait_fev.gh5"   
[35] "./studies/satsa/physical/b1_male_aehplus_gait_grip.gh5"  
[36] "./studies/satsa/physical/b1_male_aehplus_grip_fev.gh5" 
ampiccinin commented 8 years ago

@andkov - Wow! Thanks!

Quick Q – the axes seem quite different across some plots – e.g., OCTO Female gait-grip vs pek-gait.

e.g., gait: -0.1 to -0.5 in plot with grip, 10 to -20 with pek, and one outlier beyond 30.

ialsa-lab commented 8 years ago

Yes, many room for improvement is there. Keep dropping your ideas, I’ll try to implement them to the best of my ability. However, the values are what in the output. The values vary do do axes scales.

andkov commented 8 years ago

@ampiccinin , new graphs are ready that compare observed values (smoothed) and values predicted from the factor scores (fscores). View by process pair or by study.

P.S. As an idea, now any graph can be referenced by a hyperlink. I suggest we start using Issues as discussion board by providing a graph an a description of it that supports our story about it.

radc_female_aehplus_gait_grip-1

Some of the labels are mixed up: if you pay attention to scales you'll see. A few people didn't follow convention that I have proposed so we ended up with grip-gait as well as gait-grip models. I didn't really expect that go perfectly anyway, but haven't found a way to sort them out automatically. Please help me identify which graphs have labels switched, so I can correct it.

ampiccinin commented 8 years ago

Thanks @andkov. For some reason I can’t see them.

(also – would it be better if the lower left graphs were IC-SC? What was your reasoning for IP-SC? Thanks!)

andkov commented 8 years ago

@ampiccinin , are you saying that the following links ( view by process pair or by study) do not open in your browser?

Yes, you are right . That would be logical. I think i just saw some other symmetry. I agree, having IC-SC in lower left is the proper configuration. I'll correct it on the next pass

chenkaiwu commented 8 years ago

Hi Andrey,

Could you help check why gait speed against peak flow grpahs for the HRS are not available on https://github.com/IALSA/IALSA-2015-Portland/tree/master/reports/physical/fscores_scatter/figure_rmd? Thanks.