piton

Author: mtravers

src/data_visualization/violin.clj

@@ -20,6 +20,7 @@
    [hyperphor.multitool.core :as mu]
    ))
 
+
 ;; # The World's Smallest Violin (plot generating code)
 
 
@@ -44,16 +45,17 @@
     io/file
     (ImageIO/read))
 
-;;; If you want to see a full-fledged implementation of interactive violin plots for visualizaing biological data, [the BRUCE website](https://bruce.parkerici.org) has one. 
+
+;;; Violin plots are common in the scientific literature. For an example of using violin plots in a scientific domain, see the [BRUCE website](https://bruce.parkerici.org), which uses interactive violin plots to visualize data from a brain cancer research project. 
 
 ;;; ## References
 
 ;;; - [Violin Plots: A Box Plot-Density Trace Synergism](https://web.archive.org/web/20231106021405/https://quantixed.org/wp-content/uploads/2014/12/hintze_1998.pdf) Jerry L. Hintze, Ray D. Nelson
-
+;;; - [Violin Plot - Wikipedia](https://en.wikipedia.org/wiki/Violin_plot)
 
 ;;; # Data
 
-;;; We'll use this classic [dataset about penguin morphology](https://github.com/ttimbers/palmerpenguins/blob/master/README.md). <img src='man/figures/logo.png' align="right" height="138.5" /></a>. Each row describes an individual penguin, with properties like species, sex, body mass, wing size.
+;;; We'll use this classic [dataset about penguin morphology](https://github.com/ttimbers/palmerpenguins/blob/master/README.md). Each row in this dataset describes an individual penguin, with properties like species, sex, body mass, wing size.
 
 (def penguin-data-url
   "https://raw.githubusercontent.com/ttimbers/palmerpenguins/refs/heads/file-variants/inst/extdata/penguins.tsv")
@@ -64,7 +66,7 @@
 (kind/table
  (tc/random penguin-data 10))
 
-;;; # Just the points, ma'am
+;;; # Just show me the datapoints
 
 ;;; Let's start off with a simple dot-plot. We'll group the data by `species`, and turn each value for `body_mass` into a point.
 ;;; Vega just requires specifying some basic mappings (aka encodings) between data fields and visual properties. So a minimal dot plot can look like this:
@@ -79,7 +81,8 @@
       :type :nominal}}
  }
 
-;;; Vega's defaults are not always what we want, so this is the same as above with a bit of tweaking to look more like what we want. One nonobvious change: we use `:row` in place of `:y`. This is not estrictly necessary at this point, but will make it easier when we get to actual violin plots. Also, we add some randomness to (jitter) so we can better see individual points, and just for the hell of it, map another attribute (`sex`) to `:shape`.
+;;; Vega's defaults are not always what we want, so the next version has the same as structure as before, with a bit of tweaking to look more like what we want. We'll adjust the size of the graph, adjust the scale, use color.
+;;;Wwe add some randomness (jitter) so we can better see individual points, and just for the hell of it, map another attribute, `sex`, to `:shape`.
 
 ^:kind/vega-lite
 {:mark {:type "point" :tooltip {:content :data}}
@@ -105,6 +108,8 @@
  :width 800
  }
 
+;;; One nonobvious change: we use `:row` in place of `:y` for the group (species) dimension.  This is not estrictly necessary at this point, but will make it easier when we get to actual violin plots. Just to be more confusing, we reuse the `:y` encoding for the random jitter.
+
 
 ;;; # Boxplot
 
@@ -164,5 +169,5 @@
   :width 800
   })
 
-;;; That's the basics of a violin plot! In the followup page, we'll see about abstracting some of this into functions, with some variations. and we'll look at combining violin plots with dot and box plots for a richer of our data.
+;;; That's the basics of a violin plot! In the [followup page](violin2), we'll see about abstracting some of this into functions, with some variations. and we'll look at combining violin plots with dot and box plots for a richer of our data.
 

src/data_visualization/violin2.clj

@@ -21,10 +21,17 @@
    ))
 
 
-;;; # Data (repeated from first part) TODO hide
+;; # The World's Smallest Violin (plot generating code), Part 2
+
+
+;;; [Link to Part 1]()
+
+
+^:kindly/hide-code ^:kind/hidden
 (def penguin-data-url
   "https://raw.githubusercontent.com/ttimbers/palmerpenguins/refs/heads/file-variants/inst/extdata/penguins.tsv")
 
+^:kindly/hide-code ^:kind/hidden
 (def penguin-data
   (tc/dataset penguin-data-url {:key-fn keyword}))
 
@@ -36,12 +43,14 @@
 ;;; Once we know how to make a visualization, it makes sense to abstract it into a procedure, so this knowledge in a function. 
 
 
-;;; For inst
+;;; So lets do that for dot plot. We'll make a function that takes three required objects: `data`, `value-field`, and `group-field`, along with some options. 
 
 (defn dot-plot
-  [data value-field group-field]
+  [data value-field group-field
+   & {:keys [jitter? color-field]}]
   {:mark {:type "point" :tooltip {:content :data}}
    :data data
+   :transform [{:calculate "random()" :as "jitter"}]
    :encoding
    {:x {:field value-field
         :type :quantitative
@@ -50,82 +59,55 @@
           :type :nominal
           :header {:labelAngle 0 :labelAlign "left"}
           :spacing 0}
-    :color {:field group-field
+    :y {:field (if jitter? "jitter" nil)
+        :type :quantitative
+        :axis false}
+    :color {:field (or color-field group-field)
             :type :nominal
-            :legend false}  
-    }
+            :legend (if color-field true false)}}
    :height 50
    :width 800
    })
 
-;;; Which can be used like this:
-
+;;; Which can be used like this (here we'll look some differen attributes)
 
 ^:kind/vega-lite
 (dot-plot {:values (tc/rows penguin-data :as-maps)}
-          "flipper_length_mm" "year"
+          "body_mass_g" "sex"
+          :jitter? true
+          :color-field "species island"
           )
 
 
-;;; On any data set
+;;; And we can easily reuse the function on a different data set (this one is about movies)
+
 ^:kind/vega-lite
 (dot-plot {:url "https://vega.github.io/editor/data/movies.json"}
-          "US Gross" "Major Genre")
+          "US Gross" "Major Genre"
+          :jitter? true)
 
 
 
-;;; Add jitter
-
-(defn dot-plot-2
-  [data value-field group-field jitter?]
-  {:mark {:type "point" :tooltip {:content :data}}
-   :data data
-   :transform (if jitter? [{:calculate "random()" :as "jitter"}] [])
-   :encoding
-   {:x {:field value-field
-        :type :quantitative
-        :scale {:zero false}}
-    :y (when jitter?
-         {:field "jitter"
-          :type :quantitative
-          :axis false})
-    :row {:field group-field
-          :type :nominal
-          :header {:labelAngle 0 :labelAlign "left"}
-          :spacing 0}
-    :color {:field group-field
-            :type :nominal
-            :legend false}  
-    }
-   :height 50
-   :width 800
-   })
-
-
-;;; On any data set
-^:kind/vega-lite
-(dot-plot-2 {:url "https://vega.github.io/editor/data/movies.json"}
-          "US Gross" "Major Genre" true)
-
-
 ;;; # Generalize
 
-;;; This section introduces a new, and somewhat funky way of using and generalizing Vega specs.
-
-;;; Take our dot-plot abstraction above. We could parameterize it further, say :type which could be :dotplot or :boxplot. But instead, we're going to hack it by introducing a function that can merge arbitrarily nested structures. This means we can alter any aspect of the spec, at the cost of having to have some knowledge of its structure. Eg we could change the height or spacing or fonts.
+;;; This section introduces a new, and somewhat funky way of generalizing Vega specs.
 
+;;; Take our dot-plot abstraction above. We could parameterize it further, eg by adding optional arguments for height or scale or any of the many things Vega allows you to tweak.  
 
-;; TODO maybe more confuscing than it is worth here. For a later section?
+;; But instead, we're going to introduce a much more general (if somewhat unclean) way of modifying a base Vega spec – through structural merge. This makes use of a function `mu/merge-recursive` from the  (Multitool utility library)[https://github.com/hyperphor/multitool/blob/9e10c6b9cfe7f1deb496e842fc12505748a09d69/src/cljc/hyperphor/multitool/core.cljc#L1012]. This function m merges arbitrarily nested structures. This means we can alter any aspect of the spec, at the cost of having to have some knowledge of its structure. 
 
+(defn dot-plot-g
+  [data value-field group-field overrides]
+  (mu/merge-recursive
+   (dot-plot data value-field group-field false)
+   overrides))
 
-;; mu/merge-recursive is a function from the Multitool utility library [link].
 
-(defn box-plot
-  [data value-field group-field]
-  (-> (dot-plot-2 data value-field group-field false)
-      (mu/merge-recursive
-       {:mark {:type :boxplot
-               :extent :min-max}})))
+^:kind/vega-lite
+(dot-plot-g {:values (tc/rows penguin-data :as-maps)}
+            "body_mass_g" "sex"
+            {:mark {:filled true}}
+            )