A Bayesian perspective on single-shot laser characterization - Supplementing Dataset

This dataset is supporting material to the paper ''A Bayesian perspective on single-shot laser characterization" and consists of preprocessed Single-Shot FALCON data ('slopes_stacked.npy') and the corresponding timestamps ('timestamps_in_min.npy'). The dataset was obtained by analyzing raw Single-Shot FALCON images by separating the individual filter parts and extracting the local microlense foci shifts compared to a reference. In this case, the reference is an averaged image taken before the dataset, as only relative changes over time and not absolute values are shown in the paper. The shifts of the microlense foci were converted into local tilts of the wavefront. They were scaled to include the demagnification of the laser beam before imaging the near field. Lenslets without sufficient illumination, or that resulted in faulty values for other reasons, were excluded for all measurements. The structure of the set is the following: [index, lenslet position horizontal, lenslet position vertical, slope orientation, filter] index ranges from 0 to 698 and identifies each single-shot measurement. It corresponds to the entry in 'timestamps_in_min.npy'. lenslet position horizontal/vertical range from 0 to 28. For plt.imshow(), these are the dimensions that should be plotted. slope orientation is either 0 or 1. Each local wavefront tilt is expressed in two coordinates. 0 corresponds to horizontal, 1 corresponds to vertical direction. filter ranges from 0 to 8 and indicates which section, according to the filter, is considered. Indices are associated with the filters with the following central wavelength in nm in order: [763.0, 771.5, 781.1, 790.3, 804.1, 809.7, 827.4, 830.6, 842.6]. The total spectra are shown in Figure 4 of the paper. Values in the dataset are in rad. Every step in processing this dataset is described in the paper.本数据集为论文《单次激光表征的贝叶斯视角》（"A Bayesian perspective on single-shot laser characterization"）的配套支撑材料，包含预处理后的单次FALCON（Single-Shot FALCON）数据（文件名为`slopes_stacked.npy`）及其对应的时间戳数据（文件名为`timestamps_in_min.npy`）。 数据集通过分析原始单次FALCON图像得到：将图像中的各个滤光片区域分离，并提取相较于参考的子透镜（lenslet）焦点偏移量。此处的参考为数据集采集前拍摄的平均图像，因论文仅展示随时间的相对变化而非绝对数值。随后将子透镜焦点偏移量转换为波前局部倾斜量，并进行缩放以计入近场成像前激光束的缩束倍率。所有测量中均剔除了照明不足或因其他原因产生错误数值的子透镜。 数据集结构如下： [index, 子透镜水平位置, 子透镜垂直位置, 斜率取向, 滤光片编号] 各字段详细说明如下： 1. index：取值范围为0至698，用于标识单次测量，与`timestamps_in_min.npy`中的条目一一对应。 2. 子透镜水平/垂直位置：取值范围为0至28，使用`plt.imshow()`绘图时，此范围即为图像的尺寸参数。 3. 斜率取向：可取0或1。每个局部波前倾斜量可通过两个坐标分量表示，0对应水平方向，1对应垂直方向。 4. 滤光片编号：取值范围为0至8，用于标识当前对应的滤光片区域。各编号与滤光片中心波长（单位：nm）的对应关系依次为：[763.0, 771.5, 781.1, 790.3, 804.1, 809.7, 827.4, 830.6, 842.6]。论文中图4展示了完整光谱。 数据集内的数值单位为弧度（rad）。 本数据集的所有处理流程均在论文中进行了详细阐述。