Skip to content

AGRONOMY4FUTURE

Stories about crops and statistics (plus coding)

  • CURRICULUM VITAE
  • DATABASE
    • Linux
    • Structured query language (SQL)
    • EXCEL/VBA
    • Power Query
    • Data Science
  • CODING
    • Image analysis
    • R coding
    • SAS
    • PYTHON
    • Machine Learning
  • STATISTICAL MODEL
    • ANOVA
    • REGRESSION
    • SPLIT-PLOT
    • GLM
    • Mixed Model
    • REML
    • NONLINEAR
  • BASIC STAT

Image analysis

■ Object Detection by Shape

□ phenoshot() R Package: AI-Based Plant Image Analysis for Morphological Trait Measurement

■ Agronomy4future Coding Hub
□ R STUDIO: rstudio.agronomy4future.com
□ Jupyter Lab: jupyter.agronomy4future.com
□ GitHub: github.com/agronomy4future

■ Agronomy4future Data Hub
□ MySQL: mysql.agronomy4future.com
□ Cloud: cloud.agronomy4future.com
□ Stat: https://agrivoltaics.agronomy4future.com

■ Agronomy4future Research Hub
□ Paper Mining: scopusmining.agronomy4future.com
□ Research Note: note.agronomy4future.com

■ AGRONOMY4FUTURE TODAY
□ today.agronomy4future.com

JK Kim
([email protected])
"We aim to develop open-source code for agronomy"

annotate() aov () axis.title.y c(rep() case_when() colnames() data.frame() ddply () dplyr facet_grid() facet_wrap() filter() fixef() fwrmodel() gdds() geom_abline () geom_bar () geom_boxplot() geom_jitter() geom_vline() github group_by() HSD.test() ifelse() intercept0() kimindex() mutate() nls() nlsLM() normtools() pivot_longer() poly() predict() rbind() reshape2::melt() residuals() scale_fill_manual () spread() str_replace_all () subset() summarise() theme_grey() windows () write_xlsx() xtabs()

When taking pictures in the field, background noise (such as weeds and soil conditions) can make it difficult to fully detect leaves. Several software tools provide background removal. To save time, we can connect a program's API to R code to process multiple images simultaneously. Among the services I tested that offer API access, Photoroom produced the cleanest cutouts, making it the only supported provider in this workflow. You can get an API key directly from Photoroom. Please note that each processed image consumes API credits, which can be purchased on the Photoroom website.By integrating the API key into the R code, field images can be processed automatically and successfully. Of course, if the background noise is severe, it may not be 100% removed.
today.agronomy4future.com
We applied a burndown herbicide to a new research area in the agrivoltaics field. This extended area allows us to study the interactions between shading and other environmental factors to elucidate how crop yield is affected by such environments.
today.agronomy4future.com
We are currently screening lettuce varieties for shade tolerance as part of this season's agrivoltaics study. We plan to select one or two varieties to test in the actual agrivoltaics research field.
today.agronomy4future.com
When I used Zotero on Windows, I didn't think it was a good fit for me, so I stopped using it. But after switching to Linux—wow, it’s fantastic! Of course, there are some inconveniences (not being able to use MS Office is the big one), but the benefits are worth it. I'm so glad I made the switch. It's hard to describe, but everything feels so much lighter compared to Windows. # Install Zotero in Linux sudo snap install zotero-snap Delete Zotero in Linux sudo snap remove zotero-snap
today.agronomy4future.com
facetext(): Easy Text Annotation for ggplot2 Faceted Plots
In R, adding text to faceted panels can be tricky. facetext() and facetext2() make it simple using relative coordinates directly on the figures.
today.agronomy4future.com
facetext(): phenokio() R Package: Grain Size Analysis – Length, Width, and Area Metrics
When analyzing grain size, we’ve used high-throughput image scanning machines. However, if we can detect grains using R code, estimating grain size becomes possible. Recently, we developed a new R function called phenokio() specifically designed to estimate the grain area of cereals.
today.agronomy4future.com
Physiological bases of wheat grain weight response to heat waves: Post-anthesis sensitivity and responses to source-sink manipulations in contrasting cultivars
This is one of my recently published studies, conducted in Spain, on how post-anthesis #heatwaves affect grain weight in winter #wheat, and whether these effects are driven by source–sink limitations or by direct impacts on grain growth capacity.
today.agronomy4future.com
lrr(): Log Response Ratio and Effect Size Calculation
This is an R function I developed to simply calculate the log response ratio (LRR) as an effect size, which is commonly used in meta-analyses. The lrr() function automatically calculates the LRR along with its 95% confidence interval.
today.agronomy4future.com
Connecting R to Linux-Based Virtual Private Server (VPS) by DigitalOcean for Secure Data Access
Here, I share an efficient way to manage databases using a Linux-based Virtual Private Server (VPS). I don’t manage my data in Excel; instead, I primarily use SQL. I also don’t store any data on my local PC. All of my data is kept on my Linux-based server, and I access it whenever I need using a simple R function I developed, as shown below.
today.agronomy4future.com
Shading impacts on sorghum and soybean grain yields in agrivoltaics systems: Source-sink strength in response to shading
In the Midwest region of the U.S., could hashtag#agrivoltaics (AV) farming systems for sorghum and soybean be beneficial under drought conditions, as has often been reported in Mediterranean climates, by offsetting the shading effect? Additionally, if there is a negative shading effect on grain yield, which yield components are primarily driving the yield penalty, and how should farming strategies for sorghum and soybean be adjusted in AV systems? This latest paper on the #agrivoltaics study I conducted at the University of Illinois Urbana-Champaign will address these questions.
today.agronomy4future.com
agronomymap(): Spatial Heatmap Visualization for Agronomic Grid-based Field Layouts Trials
agronomymap() provides functions for generating heat maps and spatial field visualizations from agronomic trial data, designed for grid-based field layouts (row/column).
today.agronomy4future.com
gammacurve(): Estimate Gamma Distribution Parameters and Compute PDF or CDF
The Gamma distribution is ideal for modeling strictly positive, asymmetric data—unlike the Normal distribution, which can produce impossible negative values when variability is high.In this post, I demonstrate how to calculate the Gamma Probability Density Function (PDF) manually and in Excel, and introduce gammacurve(), an R package I developed to simplify creating and visualizing Gamma distributions.
today.agronomy4future.com
colorcapture(): segment and measure colored objects in images
We developed a new R function, colorcapture(). By simply switching the HSV segmentation, different colors of fruits (or grains) can be detected by R, and its surface area automatically calculated.
today.agronomy4future.com
datacume(): Compute Cumulative Summaries of Grouped Data
Just released a new R function: datacume(). This function helps you easily calculate cumulative values over time (or trials), grouped by category (also supports averaging).
today.agronomy4future.com
datacooks(): Cook's Distance Diagnostics and Outlier Detection
Just released a new R function: datacooks(). It adds diagnostic stats (predicted values, residuals, leverage, studentized residuals, Cook’s distance, etc.) directly to your model dataset — then automatically spots and flags potential outliers.
today.agronomy4future.com
nrmodel(): Quantifying Reaction Norm Plasticity from Slopes to Individual Responses
This R code is part of a series introducing R packages that present methods for calculating phenotypic plasticity. Following fwrmodel() for the Finlay-Wilkinson Regression model and deltactrl() for responsiveness, rnmodel() provides a simple way to quantify slopes as a measure of plasticity. The concept behind this code is that plasticity can be quantified by estimating slopes for individual genotypes across environments. rnmodel() offers an easy way to obtain these slopes for each specific environment.
today.agronomy4future.com
ASA meeting in 2025
At this week’s ASA meeting, I had an oral presentation on my current agrivoltaics study, where I proposed different farming strategies for sorghum and soybean at pre- and post-anthesis stages. First, the shading effect caused by solar panels results in yield penalties for both sorghum and soybean, primarily due to a decreased grain number rather than grain weight. Sorghum was more likely source-limited during grain filling, while soybean appeared to be co-limited. Intensive management (e.g., increased resources) would be necessary at the pre-anthesis stage to mitigate yield penalties, whereas post-anthesis management would be more relevant for sorghum. For soybean, preventing a lack of resources is more critical, as increasing resources post-anthesis has less impact on grain weight.
today.agronomy4future.com
SCAPES Crops Research @ University of Illinois at Urbana-Champaign
Last season, I altered the sources-sink strength of sorghum and soybean and collected valuable data. This upcoming season, I will analyze the genotypic variance of soybeans in response to shading.
today.agronomy4future.com
  • Agricultural Institute
  • Agronomy
  • AI
  • Article
  • Canada
  • Cereals
  • Conference
  • Crop breeding
  • Crop physiology
  • Data Science
  • España
  • Excel
  • Fertilizer
  • GHGs
  • Google Colab
  • hemp (Cannabis sativa)
  • JK column
  • JMP
  • Linux
  • Machine learning
  • Maize
  • Nederland
  • Principle-Centered Leadership
  • Python programming
  • R programming
  • SAS
  • Soil
  • Sorghum
  • source-sink dynamics
  • Soybean
  • SQL
  • Statistics
  • Tool
  • Uncategorized
  • United States
  • VBA
  • Wheat
  • phenoshot() R Package: AI-Based Plant Image Analysis for Morphological Trait Measurement
  • Burn down herbicide
  • Simplify Your Data Cleaning: Replace Text in R
  • sangchu: A Python Package for Green Leaf Area Detection in Greenhouse Pot Experiments
  • The Best Linear Unbiased Estimator (BLUE): Step-by-Step Guide
  • facetext() R Package: Easy Text Annotation for ggplot2 Faceted Plots
  • Database Flow Map
  • Firth’s Logistic Regression: Solving the Problem of Separation
  • Build a Cloud-Synced CLI Worklog: The Ultimate Minimalist Setup for Linux
  • How to Sync OneDrive (or Google Drive or Box) on Linux using Rclone?
  • phenokio() R Package: Grain Size Analysis – Length, Width, and Area Metrics
  • [Agronomy article] Nitrogen Cycle in Soil
  • Literature Mining for Meta-Analysis Using the scopusmining Package
  • Converting Rows to Columns in R: A Guide to Transposing Data (feat. pivot_wider and pivot_longer)
  • [R package] Log Response Ratio and Effect Size Calculation (feat. lrr)
  • Confidence interval (CI) formula for a two-sample t-test
  • Understanding Bayes’ Theorem Step by Step
  • Connecting R to Linux-Based Virtual Private Server (VPS) by DigitalOcean for Secure Data Access
  • [R package] Cook’s Distance Diagnostics and Outlier Detection (Feat. datacooks)
  • [R package] Spatial Heatmap Visualization for Agronomic Grid-based Field Layouts Trials (Feat. agronomymap)
  • Visualizing Yield Data with a Heat Map in R
  • How to analyze quadratic plateau model in R Studio?
  • What is the Gamma Distribution? Shape and Scale Parameters, and the Probability Density Function (PDF)
  • guides(fill=”none”)
  • How to Set Up RStudio Server on a Linux-Based Virtual Private Server (VPS)
  • [Data article] How to Import Data from MySQL Server to R?
  • [R package] Segment and Measure Colored Objects in Images (Feat. colorcapture)
  • [Data article] How to Import Data to a Cloud MySQL Server (DigitalOcean) Using Python from Command Prompt
  • [R package] Segment and Measure Green Objects in Images (Feat. greencapture)
  • [R package] Compute Cumulative Summaries of Grouped Data (Feat. datacume)
© 2022 - 2023 https://agronomy4future.com
© 2026 AGRONOMY4FUTURE • Built with GeneratePress