Dataset: Raw fit data of both epithelial tissues and electronic cell models

This dataset was the result of running the 3P-EIS data through the software developed for this project (https://github.com/colbyNIH/3P-EIS). Each row corresponds to a unique measurement that was run through the fitting code. The column naming conventions are:RsolA, RsolB, Ra, Ca, Rb, Cb, Rs: These names correspond to the independently verified electronic circuit element, if it was known for a particular measurementVa_rail: a built in check for voltages near the maximum range of the hardware. This was used to check for incorrect hardware configurations.chamber: "model" or "Ussing." Model corresponds to the electronic model cell and Ussing corresponds to recordings from epithelial tissues mounted in the Ussing chamber.circuitValuesKnown: A boolean used for filtering only measurements where the ground truth elements were recorded.constrain_min_and_max_f: a boolean used to indicate if a minimum and maximum frequency range was enforced for each measurement in the fitting code. The magnitude of the min and max frequency were set in the software.cross_section: the cross-sectional area of the chamber in cm2.data_location: a string used to locate the raw data for each measured file.detrendSignal: a boolean to indicate if the detrending logic was applied to the measured voltages prior to performing the STFT analysis of the signal amplitudes.filterSingals: a boolean to indicate if the measured voltage data was low-pass filtered prior to STFT analysis of the signal amplitudes.fs: the sampling frequency of the voltage data in Hz.high_RSME: a boolean to check if the voltage data was extremely noisy, indicating a poor electrode connection during recording.variables with prefid "idx_best_" the index of the best fit out of all initial guesses for a particular model.import_error: a boolean to indicate if the measured voltage data failed to import correctly.include_in_analysis: a boolean used before all raw data had been acquired to test the fitting script logic. Mainly used to avoid loading files that didn't yet exist.manually_validated: a boolean used to show which files were checked by a human to match the expected ranges of voltages.meas_ID: a unique string generated at the start of any recording. Contains date time information about recording.meas_idx: for each recording, multiple measurements could be made. This value corresponds to the nth measurement in meas_ID.n_freqs: number of frequencies applied across the system for the particular EIS measurement. normlization_method: the string used in the fitting algorithm that set the pre-processing steps on the EIS data prior to fitting.prefix "p_": Ra, Rb, Rs, etc were converted to generic suffix numbers. The prefix "p" stands for parameter and matches one of the circuit elements. In this order: RsolA, RsolB, Ra, Ca, Rb, Cb, Rs links to 1, 2, 3, 4, 5, 6, and 7. prefix "pg_": the g in this variable prefix stands for "guess." This is a guess for the parameter with the same suffix number in "p." For example, pg_absEr_1 is the guess - based on the absEr model fit - for parameter p_1 and so on. The model used in this paper (membrane ratio) is "absZr."pgmax: the maximum value for each parameter suffix used in fitting.pgmin: the minimum value for each parameter suffix used in fitting.resnormbest: the sum of all residuals for the best fit with each model after fitting.signal_buffer_end: There was a delay in the NOVA 2.1.5 software at the beginning and end of the measurement. This is a user estimated value that corrects the true start and end of each recording so that the STFT windows search the correct portion of the time domain signal for the frequency of interest. signal_buffer_start: There was a delay in the NOVA 2.1.5 software at the beginning and end of the measurement. This is a user estimated value that corrects the true start and end of each recording so that the STFT windows search the correct portion of the time domain signal for the frequency of interest. total_guesses: the total number of random starting guesses used to fit each model to the measured data. vpipette_gain_adjust: for some of the validation data, the pipette gain was set to 10x and some of the data was set to 1x. To ensure the correct amplitude was calculated after STFT analysis, the raw voltage data was divided by this value.

本数据集是将3P-EIS数据通过本项目自研软件（https://github.com/colbyNIH/3P-EIS）处理后得到的结果。每一行对应一次经过拟合代码处理的唯一测量值。列命名约定如下：
RsolA、RsolB、Ra、Ca、Rb、Cb、Rs：若某次测量已知对应电路元件，这些名称分别对应经独立验证的电子电路元件。
Va_rail：硬件电压量程上限附近的内置校验项，用于检测硬件配置错误。
chamber：取值为"model"或"Ussing"。其中"model"代表电子模型电池，"Ussing"代表安装于尤斯灌流室（Ussing chamber）的上皮组织记录样本。
circuitValuesKnown：布尔型字段，用于筛选仅保留已记录真实电路元件的测量数据。
constrain_min_and_max_f：布尔型字段，用于标识拟合代码是否为本次测量强制设置了最小与最大频率范围，频率上下限由软件预设。
cross_section：腔室的横截面积，单位为cm²。
data_location：用于定位各测量文件原始数据的字符串。
detrendSignal：布尔型字段，用于标识是否在对信号幅度进行短时傅里叶变换（Short-Time Fourier Transform, STFT）分析前，已对测量电压执行了去趋势处理。
filterSignals：布尔型字段，用于标识是否在对信号幅度进行STFT分析前，已对测量电压数据执行了低通滤波处理。
fs：电压数据的采样频率，单位为Hz。
high_RSME：布尔型校验字段，用于检测电压数据是否存在严重噪声，以判断记录过程中电极连接是否不良。
以"idx_best_"为前缀的变量：代表针对特定模型的所有初始猜测中，最优拟合结果的索引。
import_error：布尔型字段，用于标识测量电压数据是否未能正确导入。
include_in_analysis：布尔型字段，用于在所有原始数据采集完成前测试拟合脚本逻辑，主要用于避免加载尚未生成的文件。
manually_validated：布尔型字段，用于标记已由人工校验过电压范围是否符合预期的文件。
meas_ID：任意记录开始时生成的唯一字符串，包含记录的日期与时间信息。
meas_idx：单次记录可包含多次测量，该值代表本次meas_ID下的第n次测量。
n_freqs：本次电化学阻抗谱（Electrochemical Impedance Spectroscopy, EIS）测量中施加于系统的频率总数。
normalization_method：拟合算法中使用的字符串，用于设置拟合前对EIS数据的预处理步骤。
以"p_"为前缀的字段：Ra、Rb、Rs等均被转换为带通用数字后缀的形式，前缀"p_"代表"参数（parameter）"，与各电路元件一一对应。按顺序，RsolA、RsolB、Ra、Ca、Rb、Cb、Rs分别对应p_1、p_2、p_3、p_4、p_5、p_6与p_7。
以"pg_"为前缀的字段：其中"g"代表"猜测（guess）"，该字段为对应"p_"后缀编号的参数的初始猜测值。例如，pg_absEr_1为基于absEr模型拟合得到的参数p_1的初始猜测值，以此类推。本文使用的模型（膜比例模型）为"absZr"。
pgmax：拟合过程中各参数后缀对应的取值上限。
pgmin：拟合过程中各参数后缀对应的取值下限。
resnormbest：各模型完成拟合后，最优拟合结果的所有残差之和。
signal_buffer_end：NOVA 2.1.5软件在测量的起始与结束阶段存在延迟，该值为用户估算的修正参数，用于校正各记录的真实起止时间，使STFT窗口能够在时域信号的正确区间内搜索目标频率。
signal_buffer_start：同signal_buffer_end，为用户估算的记录起始修正参数。
total_guesses：为使各模型拟合至测量数据，所使用的随机初始猜测总次数。
vpipette_gain_adjust：部分验证数据的移液管增益设为10倍，其余数据设为1倍。为确保STFT分析后能够得到正确的幅度值，需将原始电压数据除以该值。