Wan-Ru Yang

I am Wan-Ru. I am a Ph.D. candidate studying GIS at UC Davis.

The transit map above summarizing my experiences in academia and data analysis. Please click the nodes for more information. The protfolio section is continuously updating.

LinkedIn Github

Remote sensing is the technology of obtaining information about an object from a distance. I used remote sensing in detecting land use change, evaluating vegetation growing status, estimating crop water usage, creating maps for my reserach area, and etc..

I have experience in analyzing remote sensing data of several sensors,
including several USGS and commercial satellite data whose properties are relevant for my research.

Spatial data packages developed by OSGeo provide handy functions for converting geospatial data to numeric data strcuture that can be analyzed using Python, R, and other programming language.

Tobler’s Law: "...the first law of geography: everything is related to everything else, but near things are more related than distant things.” (Tobler, 1970)


Data Science Immersive: A/B testing, Python statistics/machine learning packages, Regression, Ensemble Methods, Clustering, Netwrok analysis, Time Series, SVM, NLP, Neural Network, MapReduce, Spark, Visualization

Formal Courses: Quantative Geography, Biostatistics, Geostatistics, Physical Hydrology, Hydrologic Modeling, Systems Hydrology(Hydrology reserach design), Applied Multivariate Modeling

Online Courses
Coursera: Maching Learning

Applied in research projects:
exploratory data analysis; time series analysis, multinomial logistic regression, spatial panel data model, cluster analysis, non metric multidimensional scaling, and ......

Main (knowledge and daily experience): R, Python, PostgreSQL

Experienced: Javascript, D3.js, bash, c#, VBA


Corn production vs. farm acres treated with insecticide: The plot represents the total farmland area and corn growing acerage at county level during 4 agricultural survey periods.A data analysis project that comfimed the existence of evoloving pest resistance to GMO corn.
Read More (mapping with D3)
Journal Paper(pdf)


Cropping pattern in California: I applied cluster analysis and Non-metric multidimensional scaling (NMDS) on inter-annual crop variation data derived from California's pesticide use report. The purpose was to locate farmers who are more inclined to grow new crops. Left plot shows the distribution of cluster result, which has been used in developing two anlaysis that will be submitted for publication soon:
1) incorporated into an economic model to predict the effect of increasing price and adoption of sugarbeet as an energy crop
2) fitted a logistic model to examining factors that contributing to farmers' crop choice deceision.
The result indicates that spatial aggregation of data points in the same clustered group. Please check out the reuslt on google map .


Landuse mapping from remote sensing data: The images on the left shows the segmentation result of a set of composed Lansat image of my reseach area. The spectral reflectance values extreacted from AVIRIS images is used for classification with prefined labels. The training labels were created using data derived from California's pesticide report, which contains crop existence infomation at 1-square-mile squares across the entire California. This is an ongoing project planned to be done by April, 2016.


Lognormal distribution parameter estimation using method of memoent: This was a class project done in 2010, and is my very first project that completely depended on R for analysis. read more