correct vignette

Author: pglpm

vignettes/vignette_start.Rmd

@@ -903,12 +903,12 @@ Fspecies60$quantiles[, , c('5.5%', '94.5%')]
 
 
 
-## Updates in island-wise conditional frequencies: a surprising result
+## Updates in island-wise conditional frequencies
 
 We can quickly repeat the analysis above for the *conditional* frequencies on the `Biscoe` and `Dream` islands, that is, the frequencies for the two subpopulations of penguins living on those two islands. Let's directly plot the probability distributions for the two subpopulation frequencies, together with those from the previous inference.
 
-\
-For `Dream` island:
+
+### For `Dream` island
 
 
 ``` r
@@ -934,8 +934,7 @@ hist(onlyDream10, legend = 'topright', xlim = c(0, 1),
 The new frequency estimates for `Adelies` (solid red line) and `Chinstrap` (dashed blue line) are now less than 10% apart, whereas the previous estimates had them 20% apart. But this change is not unexpected: these new estimates are well within the previous credibility intervals; we were correctly warned about their possible change.
 
 
-\
-For `Biscoe` island:
+### For `Biscoe` island: a surprising result {#surprise}
 
 
 ``` r
@@ -1204,7 +1203,8 @@ Fanalysis <- Pr(Y = Y, X = X, learnt = learntall, parallel = 4)
 # 
 # Closing connections to cores.
 ```
-Here are the probability distributions for the three frequencies:
+Here are the probability distributions and the credibility intervals for the three frequencies:
+
 
 ``` r
 hist(Fanalysis, xlim = c(0, 1), col = 2:4, ## same colours as before!
@@ -1213,6 +1213,25 @@ hist(Fanalysis, xlim = c(0, 1), col = 2:4, ## same colours as before!
 
 <img src="figure/unnamed-chunk-8-1.png" width="100%" />
 
+``` r
+
+Fanalysis$values
+#            X
+# Y              Biscoe
+#   Adelie    0.2601481
+#   Chinstrap 0.0362443
+#   Gentoo    0.7036076
+
+Fanalysis$quantiles[, , c('5.5%', '94.5%')]
+#            
+# Y                5.5%     94.5%
+#   Adelie    0.2078777 0.3144158
+#   Chinstrap 0.0162262 0.0606095
+#   Gentoo    0.6453513 0.7577983
+```
+
+Conclusion:
+
 > From a sample of 344 penguins, the relative frequencies of the three species **on Biscoe island** are estimated as follows:
 > 
 > Adélie
@@ -1227,7 +1246,7 @@ hist(Fanalysis, xlim = c(0, 1), col = 2:4, ## same colours as before!
 > : estimate 0.70
 >  <br>true rel. frequency between 0.65 and 0.76 with 89% probability
 
-Note how our previous estimates from 60 samples were already quite on the spot.
+Note how our [previous estimates from 60 samples](#surprise) were already quite on the spot.
 
 
 ### **Q2b**: Adélie on three islands
@@ -1248,13 +1267,14 @@ Xvls <- 'Adelie' ## subpopulation value
 X <- data.frame(Xvls)
 colnames(X) <- Xvrt
 
+## NB: rewriting the previous 'Fanalysis' object
 Fanalysis <- Pr(Y = Y, X = X, learnt = learntall, parallel = 4)
 # 
 # Registered doParallelSNOW with 4 workers
 # 
 # Closing connections to cores.
 ```
-Here are the probability distributions for the three frequencies:
+Here are the probability distributions and confidence intervals for the three frequencies:
 
 ``` r
 hist(Fanalysis, xlim = c(0, 1), col = 5:7,
@@ -1263,6 +1283,25 @@ hist(Fanalysis, xlim = c(0, 1), col = 5:7,
 
 <img src="figure/unnamed-chunk-9-1.png" width="100%" />
 
+``` r
+
+Fanalysis$values
+#            X
+# Y             Adelie
+#   Biscoe    0.289930
+#   Dream     0.388796
+#   Torgersen 0.321274
+
+Fanalysis$quantiles[, , c('5.5%', '94.5%')]
+#            
+# Y               5.5%    94.5%
+#   Biscoe    0.232984 0.347477
+#   Dream     0.324777 0.452447
+#   Torgersen 0.264392 0.382129
+```
+
+Conclusions:
+
 > From a sample of 344 penguins, the distribution of the **Adélies** species on the three islands is estimated as follows:
 > 
 > Biscoe
@@ -1300,6 +1339,7 @@ Xvls <- sort(unique(penguin[, Xvrt])) ## subpopulation values
 X <- data.frame(Xvls)
 colnames(X) <- Xvrt
 
+## NB: rewriting the previous 'Fanalysis' object
 Fanalysis <- Pr(Y = Y, X = X, learnt = learntall, parallel = 4)
 # 
 # Registered doParallelSNOW with 4 workers