This readme file was generated on [february 2024] by [Valentín Pérez Herranz (Principal Investigator)]


___________________
GENERAL INFORMATION
___________________

Title of Dataset: Database for the paper: Deep Learning algorithm, based on convolutional
neural networks, for electrical equivalent electrical circuit recommendation for 
Electrochemical Impedance Spectroscopy.

The database consists of 544 Nyquist plots from Electrochemical Impedance Spectroscopy (EIS)
spectra classified according to their electrical equivalent circuit.


Author/Principal Investigator Information
Name: Valentín Pérez Herranz
ORCID: https://orcid.org/0000-0002-4010-0888
Institution: Universitat Politècnica de València
Address: Camino de Vera s/n
Email: vperez@iqn.upv.es

Author/Associate or Co-investigator Information
Name: Juán José Giner Sanz
ORCID: https://orcid.org/0000-0003-0441-6102
Institution: Universitat Politècnica de València
Address: Camino de Vera s/n
Email: juagisan@etsii.upv.es

Author/Associate or Co-investigator Information
Name: Fermín Sáez Pardo
ORCID: https://orcid.org/0009-0000-0759-6781
Institution: Universitat Politècnica de València
Address: Camino de Vera s/n
Email: fersaepa@etsii.upv.es

Date of data collection: 2024 

Geographic location of data collection: Valencia, Spain

Keywords: Artificial Vision, Convolutional Neural Networks, Deep Learning, Electrical Equivalent Circuits, Electrochemical Impedance Spectroscopy.

Information about funding sources that supported the collection of the data: Espectroscopía
de impedancias electroquímicas. Identificación de circuitos eléctricos equivalentes vía
inteligencia artificial project. Project founded by Universidad Politécnica de Valencia (PAID 01-22)


__________________________
SHARING/ACCESS INFORMATION
__________________________

Licenses/restrictions placed on the data: COPYRIGHT

Links to publications that cite or use the data:



______________________
DATA & FILE OVERVIEW
______________________

File List:    

- Database: A .csv file with 544 EIS spectra, in which each Nyquist plot is represented by 160 numerical values.

- Image database: A .zip file that consists of 5 folders (Type_1, Type_2, Type_3, Type_4, and Type_5). 
Each folder stores the EIS spectra related to an specific type of equivalent electrical circuit; 
and the EIS spectra are in image format (.png) with a 100x100 pixel resolution.

METHODOLOGICAL INFORMATION


Database:

The database was generated from the original database generated by Shan Zu and coworkers. 
Shan Zu and coworkers generated the original database by digitalizing Nyquist plots obtained 
from scientific papers. The digitization process consisted in obtaining 80 points from each 
Nyquist plot using the online tool WebPlotDigitizer. The original database has some problems: 
noise, disorder, and unmatching classifications between the original scientific publication 
and the original database. 

In our database, the aforementioned problems were solved by re-digitalizing (i.e. carrying 
out again the digitalization process using the WebPlotDigitizer tool) the EIS spectra with
noise; ordering the numerical values from the EIS spectra with disorder; and correctly 
classifying the EIS spectra with unmatching classifications.

Image database:

The image database was generated from the data in the database. This process was carried out
in Python 3.10.11, generating 544 images (each image with square axis and removing the axis
representation) ordered in different folders according to their label.