Projects

Please view the full write ups for each portfolio item.

Portfolio

This page features descriptions of projects that use data science techniques or interesting analyses. These analyses are not yet published, and therefore, the data and code are not publically available. Projects described here use Python to gather information from APIs, automate processes, and gather data. Data analyses, cleaning, and predictions were made using R. In both languages, I have written custom functions, created outputs for data visualization, and used a variety of external packages to complete tasks that align with hypotheses. All of these projects are ecological and entomological in nature, however, the concepts have broad implications across disciplines.

Japanese beetle project

We collected Japanese beetles every two weeks at 44 sites across Twin Cities, Minnesota. To help describe and control for variability, I gathered weather and remote sensing data using Python to communicate with various APIs. R and Python were used interchangably throughout, due to strengths and weaknesses of each.

Field and spore (microscopy and DNA) data were collected collaboratively with lab members. Data mining and analysis shown here were independent.

  • Japanese beetle project
  • Age cohort analysis

    Honey bees of specific ages in a colony have defined jobs, that change over the course of an individual bee's life. Worker bees with certain jobs interact with food resources differently. For instance, a nurse bee (between 4 and 12 days old) uses bee bread as food for larvae. Forager bees (22 to 42 days) collect pollen and nectar from plants. In this experiment, researchers were interested in the age at which bees consumed pollen patties, a supplemental food source given to colonies by beekeepers. In this analysis, we used model selection techniques, controlled for inconsistencies, utilized weather data, created a temporal autocorrelation structure, and accounted for background mortality in bees.

    Data were collected by Emily Noordyke as part of a master's thesis, advised by James D. Ellis. Analysis was conducted collaboratively by Kaylin Kleckner (PhD student at the University of Florida) and Cody Prouty.

  • Honey bee age cohort analysis