How to download and use.
- download the release zip
- install the required libraries; see requirements.txt. you will need numpy, plotly, openpyxl, pandas, pyparsing, scipy, matplotlib
- make environment and pip intall streamlit therein
- set the path for pykeri in the root of release
- command "streamlit run streamlit_app.py"
Ultrahigh quality thermoelctric db This teMatDb aims ultrahigh self-consistent transport properties for machine learning and transport mechanism analysis. Thermoelectric properties of each sample can be browsed. Using self-consistency criteria for database, one can list up the erroneous samples using the data filters and obtain the errorless sample lists also.
Here we have three data filters.
- average ZT filter for average ZT deviation between ZT direictly digitized from the figure and ZT calculated from TEP curves. This is sensitive to T range mismatch and ZT bias error.
- peak ZT filter for peak ZT deviation between figure and TEP. This is sensitive to peak ZT bias error and unwanted extrapolation, mainly cuased by data digitization resolution.
- ZT interpolation error for the intersected temperature range. This is sensitive to poor interpolation, maingly caused by phase transition and exponential behavior.