This is not a very exciting post!

The lilies

Our front yard has a small patch of day lilies. Several years ago, I started to count the number of lilies open each day. I missed some days, and in 2024 we were not at home after June 30.

Counts

I kept count of number of lilies flowering each day in a simple text file. Each line represents a year. The line starts with the first day of flowering. Subsequent numbers represent the counts on successive days; NA is used for counts on days I missed. I used simple editor commands to transform this into a JSON array of objects. Each object represents a year, with start and count fields. JavaScript transforms this object into columns displayed in the table below.

Date	Count

Visualization

Properties of flowering phenology are conveniently summarized by a plot of cumulative count versus date. Flowering started around June 25 in 2022, 2023, and 2025; for some reason flowering in 2024 started more than a week earlier (June 17). Daily counts increased to a more-or-less constant rate (maybe weakly bimodal?). The total number of flowers in 2025 (530) was noticeably greater than the number in other years (320 - 370). This was due to production of more flowers per day, rather than to a longer flowering period.

Lines are colored using the ‘Hue / Saturation / Lightness’ (HSL) scale. All lines have the same hue and saturation; years progress from lighter to darker values.

Notes

I prototyped this in R, where I am very proficient. In JavaScript, the JSON data were transformed to a more normalized form and displayed using the DataTables library. Plots were done using Plotly.js. Google Search Gemini AI helped me through various programming obstacles.

I used a JavaScript ES module to isolate the code from text. Gemini pointed me to an importmap to help import DataTables; plots are a simple import of the Plotly.js JavaScript library.