Dataset used in the publication entitled "Application of machine learning to assess the influence of microstructure on twin nucleation in Mg alloys"

Documentation for the Dataset used in the publication entitled "Application of machine learning to assess the influence of microstructure on twin nucleation in Mg alloys" ** These datasets comprise the 2D EBSD data acquired in the Mg-1Al (at.%) alloy and AZ31 Mg alloy, analyzed with MTEX 7.0 software. **** More details about the experimental techniques can be found in the publication "Biaobiao Yang, Valentin Vassilev-Galindo, Javier Llorca, Application of machine learning to assess the influence of microstructure on twin nucleation in Mg alloys. npj Computational Materials, 2024." **

1. AZ31_ML.xlsx- Description: Both twin and grain data were acquired by EBSD from AZ31 Mg sample before and after deformation at the same area- Number of grains: 2640 (rows == grains) corresponding to three samples deformed in different orientations: S0, S45, and S90- Number of analyzed variables (features): 31 (columns == grain characteristics)

- Variable description by columns:        1- (Twinned) - type: boolean        Description: Indicates if the grain twinned or not after deformation                0: non-twinned grain                1: twinned grain        2- (Orientation) - type: numerical (integer)        Description: The loading (tensile) direction with respect to the c axis of lattice        3- (Strain_level) - type: numerical (float)        Description: The maximum strain level after deformation         4- (Grain_size) - type: numerical (float)        Description: The equivalent circle diameter (in micrometers) of the grain before deformation.        5- (Triple_points) - type: numerical (integer)        Description: The number of triple points of the grain before deformation         6- (Near_edge) - type: boolean        Description: Indicates if the grain is located near the edge of the 2D EBSD or not. This feature was used to filter out from the final dataset the grains near the edge of the sample. Hence, only those entries with Near_edge value of 0 were used to train and test the machine learning models.                 0: not near the EBSD edge                1: near the EBSD edge        7-12- (T_SF*) - type: numerical (float)        Description: The twinning Schmid factor based on the loading condition, orientation of parent grain and twin variants information.                T_SF1: The highest Schmid factor of extension twinning                T_SF2: The 2nd highest ...                 T_SF3: 3rd                T_SF4: 4th                T_SF5: 5th                T_SF6: The lowest Schmid factor of extension twinning        13-15- (S_SF*) - type: numerical (float)        Description: The Schmid factor for basal slip based on the loading condition, orientation of parent grain, and slip system information. Only the basal slip system is considered because it is the dominant deformation slip system in Mg during deformation.                S_SF1: The highest Schmid factor of basal slip                S_SF2: The second highest or the middle Schmid factor of basal slip                 S_SF3: The lowest Schmid factor of basal slip        16- (Neighbor_grain_n) - type: numerical (integer)        Description: The number of neighbors of the grain before deformation.         17-19- (B-b_m) - type: numerical (float)        Description: The Luster-Morris geometric compatibility factor (m') between the basal slip systems of the grain and its neighbors. Although there are 3 possible basal slip systems, only the one with the highest Schmid factor was considered to compute m′. Only maximum, minimum, and mean values were included in the dataset.                (Max_B-b_m): The highest basal - basal m' between the grain and its neighbors                 (Min_B-b_m): The lowest basal - basal m' between the grain and its neighbors                 (Mean_B-b_m): The average basal - basal m' between the grain and its neighbors        20-22- (B-t_m) - type: numerical (float)        Description: The Luster-Morris geometric compatibility factor (m') between the 6 extension twin variants of the grain and the basal slip systems of its neighbors. Although there are 3 possible basal slip systems, only the one with the highest Schmid factor was considered to compute m'. However, all 6 twinning variants have been considered, given that slip induced twinning is a localized process. Only maximum, minimum, and mean values were included in the dataset.                (Max_B-t_m): The highest basal - twin m' between the grain and its neighbors                 (Min_B-t_m): The lowest basal - twin m' between the grain and its neighbors                (Mean_B-t_m): The average basal - twin m' between the grain and its neighbors        23-25- (GB_misang) - type: numerical (float)        Description: The misorientation angle (in º) between the grain and its neighbors. In fact, disorientation angle is used for the misorientation angle. Only maximum, minimum, and mean values were included in the dataset.                (Max_GBmisang): The highest GB misorientation angle between the grain and its neighbors                 (Min_GBmisang): The lowest GB misorientation angle between the grain and its neighbors                (Mean_GBmisang): The average GB misorientation angle between the grain and its neighbors        26-28- (delta_Gs) - type: numerical (float)        Description: Grain size difference (in micrometers) between a given grain and its neighbors. The grain size is calculated as the diameter of a circular grain with the same area of the grain. Only maximum, minimum, and mean values were included in the dataset.                (Max_deltaGs): The highest grain size difference between the grain and its neighbors                (Min_deltaGs): The smallest grain size difference between the grain and its neighbors                (Mean_deltaGs): The average grain size difference between the grain and its neighbors        29-31- (delta_BSF) - type: numerical (float)        Description: The difference in the basal slip Schmid factor between a given grain and its neighbors. Only the highest basal slip Schmid factor is considered. Only maximum, minimum, and mean values were included in the dataset.                (Max_deltaBSF): The highest basal SF difference between the grain and its neighbors                (Min_deltaBSF): The smallest basal SF difference between the grain and its neighbors                 (Mean_deltaBSF): The average basal SF difference between the grain and its neighbors

2. Mg1Al_ML.xlsx- Description: Both twin and grain data were acquired by EBSD from Mg-1Al (at.%) sample before and after deformation at the same area- Number of grains: 1496 (rows == grains) corresponding to two true strain levels: ~6%, and ~10%.- Number of analyzed variables (features): 31 (columns == grain characteristics)

- Variable descriptions by columns are the same as those of AZ31_ML.xlsx