Macroexpand 2025 Noj: Civitas Workshop

Author

Affiliation

Timothy Pratley

Hummi

^{:clay {:title  "Macroexpand 2025 Noj: Civitas Workshop"
         :quarto {:author      :timothypratley
                  :image       "macroexpand_workshop_tableplot.png"
                  :type        :post
                  :date        "2025-10-17"
                  :category    :clay
                  :tags        [:clay :workflow]}}}
(ns scicloj.tableplot.ideas.macroexpand-workshop-tableplot
  (:require [tablecloth.api :as tc]
            [scicloj.tableplot.v1.plotly :as tp]))

We built this code in the Civitas Workshop. We selected Tableplot as the topic to write about.

Everything about Tableplot is inspired by R, there everything is about tables - dataframes. We hope to have our own version of that.

It is a limitation at the moment to require a dataset.

(def scatter-ds
  (tc/dataset {:x [1 2 3 4 5]
               :y [10 20 15 25 18]
               :z [1 2 1 2 1]}))

We call tableplot on the dataset:

(-> scatter-ds
    (tp/base {:=title "Sample Scatter Plot"})
    (tp/layer-point {:=x :x
                     :=y :y}))

Why is the plot not Kindly annotated? Tableplot returns annotated values.

(-> scatter-ds
    (tp/base {:=title "Sample Scatter Plot"})
    (tp/layer-point {:=x :x
                     :=y :y})
    (meta))

#:kindly{:kind :kind/fn, :options nil}

The kind is :kind/fn which is a transformation. Where is the function?

(-> scatter-ds
    (tp/base {:=title "Sample Scatter Plot"})
    (tp/layer-point {:=x :x
                     :=y :y})
    (:kindly/f))

#'scicloj.tableplot.v1.plotly/plotly-xform

hidden in the value is the :kindly/f.

Other types of plots:

Using layer-bar instead of layer-point

(-> scatter-ds
    (tp/base {:=title "Sample Scatter Plot"})
    (tp/layer-bar {:=x :x
                   :=y :y}))

We can plot different columns of the dataset as layers.

(-> scatter-ds
    (tp/base {:=title "Sample Scatter Plot"})
    (tp/layer-line {:=x :x
                    :=y :z})
    (tp/layer-bar {:=x :x
                   :=y :y}))

You don’t need a big idea to write a Civitas post, small explorations are welcome.

source: src/scicloj/tableplot/ideas/macroexpand_workshop_tableplot.clj