Skip to content

Commit

Permalink
ran similarity analysis in verbose mode
Browse files Browse the repository at this point in the history
  • Loading branch information
lbechberger committed May 27, 2022
1 parent c1f28dd commit 35629ab
Show file tree
Hide file tree
Showing 2 changed files with 175 additions and 2 deletions.
2 changes: 1 addition & 1 deletion code/shell_scripts/Shapes/data_analysis.sh
View file
Original file line number Diff line number Diff line change
Expand Up @@ -114,7 +114,7 @@ done

echo ' comparing visual and conceptual similarity'
# first: statistical analysis
Rscript code/mds/data_analysis/analyze_visual_conceptual.r -c data/Shapes/mds/data_set/individual/similarities/conceptual_10.csv -v data/Shapes/mds/data_set/individual/similarities/visual_10.csv &> data/Shapes/mds/similarities/rating_type/statistical_analysis.txt
Rscript code/mds/data_analysis/analyze_visual_conceptual.r -c data/Shapes/mds/data_set/individual/similarities/conceptual_10.csv -v data/Shapes/mds/data_set/individual/similarities/visual_10.csv --verbose &> data/Shapes/mds/similarities/rating_type/statistical_analysis.txt
# second: finding item pairs with large and small differences
python -m code.mds.data_analysis.find_item_pair_differences data/Shapes/mds/similarities/rating_type/visual/aggregated_ratings.pickle data/Shapes/mds/similarities/rating_type/conceptual/aggregated_ratings.pickle &> data/Shapes/mds/similarities/rating_type/differences.txt

Expand Down
175 changes: 174 additions & 1 deletion data/Shapes/mds/similarities/rating_type/statistical_analysis.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,49 @@ Loading required package: ordinal
[1] "CLMM analysis on full data"
[1] "--------------------------"
[1] " Fitting alternative hypothesis (main effect of ratingType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ ratingType + (1 + ratingType | pairID)
data: data

link threshold nobs logLik AIC niter max.grad
logit flexible 35400 -36011.58 72039.16 1045(9379) 4.54e-03

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 4.541 2.131
ratingTypevisual 2.642 1.625 -0.844
Number of groups: pairID 1770

Coefficients:
ratingTypevisual
0.6328

Thresholds:
1|2 2|3 3|4 4|5
0.9563 1.9028 2.8714 4.3865
[1] ""
[1] " Fitting null hypothesis (no effect of ratingType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ 1 + (1 + ratingType | pairID)
data: data

link threshold nobs logLik AIC niter max.grad
logit flexible 35400 -36100.74 72215.49 1105(10653) 7.32e-02

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 4.573 2.138
ratingTypevisual 2.787 1.669 -0.835
Number of groups: pairID 1770

No Coefficients

Thresholds:
1|2 2|3 3|4 4|5
0.3651 1.3126 2.2864 3.8143
[1] ""
[1] " Likelihood ratio test..."
Likelihood ratio tests of cumulative link models:

Expand All @@ -23,7 +65,49 @@ Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’
[1] "CLMM analysis on between-category data"
[1] "--------------------------------------"
[1] " Fitting alternative hypothesis (main effect of ratingType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ ratingType + (1 + ratingType | pairID)
data: between

link threshold nobs logLik AIC niter max.grad
logit flexible 33000 -32434.29 64884.57 1063(10161) 7.63e-02

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 2.399 1.549
ratingTypevisual 2.083 1.443 -0.842
Number of groups: pairID 1650

Coefficients:
ratingTypevisual
0.769

Thresholds:
1|2 2|3 3|4 4|5
1.244 2.210 3.227 5.047
[1] ""
[1] " Fitting null hypothesis (no effect of ratingType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ 1 + (1 + ratingType | pairID)
data: between

link threshold nobs logLik AIC niter max.grad
logit flexible 33000 -32575.26 65164.52 953(10474) 4.69e-02

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 2.394 1.547
ratingTypevisual 2.404 1.550 -0.829
Number of groups: pairID 1650

No Coefficients

Thresholds:
1|2 2|3 3|4 4|5
0.6669 1.6354 2.6589 4.4907
[1] ""
[1] " Likelihood ratio test..."
Likelihood ratio tests of cumulative link models:

Expand All @@ -43,7 +127,53 @@ Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’
[1] "CLMM analysis on within-category data"
[1] "--------------------------------------"
[1] " Fitting alternative hypothesis (interaction betwen ratingType and visualType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ ratingType * visualType + (1 + ratingType | pairID)
data: within

link threshold nobs logLik AIC niter max.grad
logit flexible 2400 -2725.27 5470.55 1357(4085) 5.53e-03

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 0.5850 0.7648
ratingTypevisual 0.3418 0.5847 -0.682
Number of groups: pairID 120

Coefficients:
ratingTypevisual visualTypeVV
-0.9774 -0.7530
ratingTypevisual:visualTypeVV
-1.2880

Thresholds:
1|2 2|3 3|4 4|5
-3.467 -2.995 -2.378 -1.403
[1] ""
[1] " Fitting null hypothesis (no interaction of ratingType and visualType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ ratingType + visualType + (1 + ratingType | pairID)
data: within

link threshold nobs logLik AIC niter max.grad
logit flexible 2400 -2741.45 5500.90 912(3540) 4.68e-03

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 0.8781 0.9371
ratingTypevisual 0.7742 0.8799 -0.812
Number of groups: pairID 120

Coefficients:
ratingTypevisual visualTypeVV
-1.724 -1.707

Thresholds:
1|2 2|3 3|4 4|5
-4.018 -3.551 -2.944 -1.977
[1] ""
[1] " Likelihood ratio test..."
Likelihood ratio tests of cumulative link models:

Expand All @@ -61,6 +191,28 @@ model.within 10 5470.5 -2725.3 32.357 1 1.283e-08 ***
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] ""
[1] " Fitting null hypothesis (no effect of ratingType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ visualType + (1 + ratingType | pairID)
data: within

link threshold nobs logLik AIC niter max.grad
logit flexible 2400 -2801.32 5618.63 791(3026) 1.34e-02

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 1.952 1.397
ratingTypevisual 3.755 1.938 -0.909
Number of groups: pairID 120

Coefficients:
visualTypeVV
-1.642

Thresholds:
1|2 2|3 3|4 4|5
-2.8983 -2.4252 -1.8119 -0.8476
[1] ""
[1] " Likelihood ratio test..."
Likelihood ratio tests of cumulative link models:

Expand All @@ -78,6 +230,28 @@ model.within 10 5470.5 -2725.3 152.09 2 < 2.2e-16 ***
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] ""
[1] " Fitting null hypothesis (no effect of visualType)"
Cumulative Link Mixed Model fitted with the Laplace approximation

formula: ratings ~ ratingType + (1 + ratingType | pairID)
data: within

link threshold nobs logLik AIC niter max.grad
logit flexible 2400 -2783.37 5582.74 878(3516) 1.04e-03

Random effects:
Groups Name Variance Std.Dev. Corr
pairID (Intercept) 0.7262 0.8522
ratingTypevisual 0.8102 0.9001 -0.066
Number of groups: pairID 120

Coefficients:
ratingTypevisual
-1.62

Thresholds:
1|2 2|3 3|4 4|5
-3.087 -2.612 -1.996 -1.027
[1] ""
[1] " Likelihood ratio test..."
Likelihood ratio tests of cumulative link models:

Expand All @@ -93,4 +267,3 @@ model.within.null.visualType 8 5582.7 -2783.4
model.within 10 5470.5 -2725.3 116.19 2 < 2.2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

0 comments on commit 35629ab

Please sign in to comment.