This readme/codebook.txt file was generated on 2016/05/18 by Kristina Golubiewski-Davis
-------------------
GENERAL INFORMATION
-------------------
1. Title of Dataset
Supplementary Data for Reconstructing Past Craft Networks: A case study using 3D scans of Late Bronze Age swords to reconstruct specialized craft networks
2. Author Information
Name: Kristina Golubiewski-Davis
Institution: University of Minnesota
Email: golu0011@umn.edu
3. Date of data collection: 2014/09/01 - 2016/13/15
4. Geographic location of data collection (where was data collected?):
3D scans from which the data was collected were scanned in Germany, Austria, Hungary, and Croatia.
5. Information about funding sources that supported the collection of the data:
A. Summer Research Grant in Austrian/Central European Studies from the University of Minnesota Austrian Studies
B. Wenner-Gren Dissertation Fieldwork Grant
C. Hella Mears Summer Fellowship
--------------------------
SHARING/ACCESS INFORMATION
--------------------------
1. Licenses/restrictions placed on the data: None
2. Publication(s) that cite or use the data:
Golubiewski-Davis, Kristina. 2016. Reconstructing Past Craft Networks: A case study using 3D scans of Late Bronze Age swords to reconstruct specialized craft networks. PhD dissertation, University of Minnesota (publication in process).
3. Was data derived from another source? Yes
Some of the data was derived from images found in the Praehistoriche Bronzefunde series.
A. Kemencezei, Tibor. 1991. Die Schwerter in Ungarn II. Vol. Prähistorische Bronzefunde IV Band 9. München: C.H. Beck'sche Verlagsbuchhandlung.
B. Krämer, Walter. 1985. Die Vollgriffschwerter in Österreich und der Schweiz. Vol. Prähistorische Bronzefunde IV Band 10. München: C.H. Beck'sche Verlagsbuchhandlung.
C. Von Quillfeldt, Ingeborg 1995. Die Vollgriffschwerter in Süddeutschland. Vol. Prähistorische Bronzefunde IV Band 11. Stuttgart: Franz Steiner Verlag.
D. Harding, Anthony. 1995. Die Schwerter im ehemaligen Jugoslawien. Vol. Prähistorische Bronzefunde IV Band 14. Stuttgart: Franz Steiner Verlag.
E. Wüstemann, H. 2004. Die Schwerter in Ostdeutschland. Vol. Prähistorische Bronzefunde IV Band 15. Stuttgart: Franz Steiner Verlag.
F. Friedrich Laux. 2009. Die Schwerter in Niedersachsen. Vol. Prähistorische Bronzefunde IV Band 17. Stuttgart: Franz Steiner Verlag.
4. Recommended citation for the data:
Golubiewski-Davis, Kristina M. (2016). Supplementary Data for Reconstructing Past Craft Networks: A case study using 3D scans of Late Bronze Age swords to reconstruct specialized craft networks, PhD. Dissertation. Retrieved from the University of Minnesota Digital Conservancy, http://dx.doi.org/10.13020/D6PK5C.
---------------------
DATA & FILE OVERVIEW
---------------------
File List:
A. [Sword_Profile_Images.zip] :
This folder contains the black and white profile outlines for each of the blades.
There are separate folders for blade cross sections, blade profiles, hilt side profiles, and hilt top profiles.
Each image is saved in bitmap format, which is the necessary format to extract the fourier transforms using SHAPE V1.3.
B. [Fourier_Transform_Files.zip] :
These folders contain .nef files. They were created using SHAPE V1.3.
There are separate folders for the blade profile, the cross section of the blade, and the top and side profiles of the hilt.
For each type of sword profile (blade, blade cross section, side hilt, and top hilt), the corresponding folder contains individual files for each group within the profile type, a file for the average of groups, and a file that inludes all the groups.
For example, there were 10 hilt groups in the study. Therefore, the folder [Hilt_Top_Profile_Fourier_Transformthere] contains the files for each group (1-10), a file for the average of each of the ten groups and a file that has the top hilt profiles of all hilts.
C. [Network_Matrices.zip] :
This folder contains .csv files.
Blade Node Data contains a spreadsheet of the groups to which each blade belongs. Comma separated.
The remaining files contain matrix data intended to be read by Gephi. These matrix files are listed below.
[DecorativeMatrix.csv] : links based on shared decorative groups. The maximum possible link weight is 6. Semi-colon separated.
[IndividualBladeHilt.csv] : links based on shared hilt and blade groups. Weight can be 0 (no link), 1 (either a blade or a hilt link), or 2 (both a blade and hilt link). Semi-colon separated.
[IndividualBladeHiltDecoration.csv] : Combination of the previous two matrices. The maximum possible link weight is 8. Semi-colon separated.
[MinSpanLinks.csv] : This is a list of links by group and their weights, as determined by the methodology described in chapter 11 of the dissertation.
D. [Statistics.zip] :
This folder contains the code and data used in the statistical analysis. The files are listed below.
[LocationAndClusterData.csv] : a spreadsheet of the clusters each blade belongs to.
[Sword_Statistics.sas] : the sas code for the project.
[Sword_Statistics.txt] : the sas code for the project saved in .txt format.
[SwordDataStats.csv] : all of the data collected on each sword for the project.
2. Are there multiple versions of the dataset? No
--------------------------
METHODOLOGICAL INFORMATION
--------------------------
1. Description of methods used for generating and processing the data:
Refer to- Golubiewski-Davis, Kristina. 2016. Reconstructing Past Craft Networks: A case study using 3D scans of Late Bronze Age swords to reconstruct specialized craft networks. PhD dissertation, University of Minnesota (publication in process).
2. Instrument- or software-specific information needed to interpret the data:
The .bmp files in [Sword_Profile_Images.zip] can be imported into SHAPE V1.3 to extract the fourier transforms. At the time of upload, SHAPE V1.3 is a free program available to download here: http://lbm.ab.a.u-tokyo.ac.jp/~iwata/shape/
The .nef files in [Fourier_Transform_Files.zip] were created using SHAPE V1.3. They can be read as .txt files to view the harmonics associated with each blade. When read by the program NefViewer in the SHAPE suite, you can view the fourier transform for each blade. The files can also be imported into the PrinComp progrm by SHAPE to determine principle components for the fourier transforms.
The .csv matrix files in [Network_Matrices.zip] can be read by Gelphi, a free network visualization software. At the time of upload, it is available to download here: https://gephi.org/. The matrices were created by hand in excel after a series of principle component analysis, cluster analysis, ANOVA, and minimum spanning trees were performed in the SAS 9.4 software. [minspanLinks.csv] was based on the minimum spanning tree data. [DecorativeMatrix.csv], [IndividualBladeHiltDecoration.csv], and [IndividualBladeHilt.csv] were based on cluster analysis data. Matrices must be semi-colon separated to be read properly by Gephi. To visualize the matrices in different ways, open the file in GEPHI. All graphs are undirected.Then go to the "data laboratory" tab and import BladeNodeData.csv to the nodes table. Choose whichever columns you'd like to compare. If you are using Geo Layout with Gephi 0.8.2, you will need to import the Lat and Lon as type "double".
Statistics for the project were run in SAS 9.4 software. [Statistics.zip]
-----------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: [Network_Matrices.zip/BladeNodeData.csv]
-----------------------------------------------------------------------
All cluster information was determined using PROC ACECLUS in SAS 9.4 software as described in chapter 9 of the dissertation.
Number of variables: 12
Number of cases/rows: 111
Variable List:
A. ID: The ID number of the blade. The number before the blade indicates which volume of the Praehistorische Bronzefunde (PBF) the blade came from followed by the idenfication number of the blade in that book.
Ex. 00.001 - The sword is not found in any PBF volume.
Ex. 09.067 - The sword can be found in PBF volume 9 under the published number 67.
B. Style: The highest level typological style assigned by the PFB series.
C. Lat: The latitude of the find site expressed in degree format.
D. Lon: The longitude of the find site expressed in degree format.
E. Loc. Est.: Whether or not the find site latitude and longitude were estimated based on shape. The process for this estimation can be found in the "Other Analysis" subsection of Chapter 9.
F. LocCluster6: The cluster to which the sword belongs based on location. There are 6 location clusters.
G. BladesCluster12: The cluster to which the sword belongs based on blade profile. There are 12 blade profile clusters.
H. HiltCluster10: The cluster to which the sword belongs based on hilt profile. There are 10 hilt profile clusters.
I. CircleClus: The cluster to which the sword belongs based on the shape of decorative circles on the hilt. There are 3 circle clusters.
J. DashClus: The cluster to which the sword belongs based on the shape of dashed lines on the hilt. There are 5 dash clusters.
K. CurveClus: The cluster to which the sword belongs based on the shape of curved parallel lines along the shoulder of the hilt. There are 4 curved parallel line clusters.
L. StraightClus: The cluster to which the sword belongs based on the shape of straight parallel lines along the body of the hilt. There are 8 straight parallel clusters.
--------------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: [Statistics.zip/LocationAndClusterData.csv]
--------------------------------------------------------------------------
All cluster information was determined using PROC ACECLUS in SAS 9.4 software as described in chapter 9 of the dissertation.
Number of variables: 15
Number of cases/rows: 111
Variable List:
A. Number: The ID number of the blade. The number before the blade indicates which volume of the Praehistorische Bronzefunde (PBF) the blade came from followed by the idenfication number of the blade in that book.
Ex. 00.001 - The sword is not found in any PBF volume. Ex. 09.067 - The sword can be found in PBF volume 9 under the published number 67.
B. Lat: The latitude of the find site expressed in degree format.
C. Lon: The longitude of the find site expressed in degree format.
D. Loc.Est.: Whether or not the find site latitude and longitude were estimated based on shape. The process for this estimation can be found in the "Other Analysis" subsection of Chapter 9.
E. Style: The highest level typological style assigned by the PFB series.
F. Find: The type of site the sword was found at.
G. LocCluster6: The cluster to which the sword belongs based on location. There are 6 location clusters.
H. BladesCluster12: The cluster to which the sword belongs based on blade profile. There are 12 blade profile clusters.
I. HiltCluster10: The cluster to which the sword belongs based on hilt profile. There are 10 hilt profile clusters.
J. CrossSectionCluster12: The cluster to which the sword belongs based on the the blade cross section profile. There are 12 cross section profile clusters.
K. RivetsCluster15: The cluster to which the sword belongs based on rivet profile. There are 15 rivets clusters.
L. CircleClus: The cluster to which the sword belongs based on the shape of decorative circles on the hilt. There are 3 circle clusters.
M. DashClus: The cluster to which the sword belongs based on the shape of dashed lines on the hilt. There are 5 dash clusters.
N. CurveClus: The cluster to which the sword belongs based on the shape of curved parallel lines along the shoulder of the hilt. There are 4 curved parallel line clusters.
O. StraightClus: The cluster to which the sword belongs based on the shape of straight parallel lines along the body of the hilt. There are 8 straight parallel clusters.
------------------------------------------------------------------
DATA-SPECIFIC INFORMATION FOR: [Statistics.zip/SwordDataStats.csv]
------------------------------------------------------------------
Chapter 8 and Appendix A of the dissertation contain detailed explanations of the data collection procedures and protocols, including pictures.
Number of variables: 201
Number of cases/rows: 111
Variable List:
A. Number: In this study, the blades identification number is comprised of the Praehistorische Bronzefunde (PBF) volume number followed by the blade number in that volume. The series number is IV for all of the volumes related to sword in the PBF. For example, blade 09.005 can be found in PBF series IV volume 9 under the blade number 5.
B. Style: The typological designation from the PBF.
C. Lat: Latitude of the find site, as extrapolated from My Nasa Data (mynasadata.larc.nasa.gov).
D. Lon: Longitude of the find site, as extrapolated from My Nasa Data (mynasadata.larc.nasa.gov).
E. Find: The type of site the blade was found at.
F. Length: The length of the blade, as published in the PBF.
G-H. R1x, R1y, R1z: The x,y,z position of the rivet designated R1, extrapolated in Geomagic using the extract reference point feature. (http://www.geomagic.com/en/)
J-L. R2x, R2y, R2z: The x,y,z position of the rivet designated R2, extrapolated in Geomagic using the extract reference point feature. (http://www.geomagic.com/en/)
M. TangHeight: The height of the tang at the point where the blade and hilt meet along the center line of the sword.
N-O. CircleDist, StDev_Circle: The average distance between the centers of concentric circles on a hilt and the standard deviation of that average.
P-Q. Radius, StDev_Radius: The average radius of the largest circle in a set of concentric circles on a hilt, and the standard deviation of that average.
R-S. CircleLine, StDev_CircleLine: The average distance between the circles with a set of concentric circles on a hilt, and the standard deviation of that average.
T-U. CurvedDist, StDev_CurvedDist: The average distance between curved parallel lines on a hilt, and the standard deviation of that average.
V-W: StraightDist, StDev_StraightDist: The average distance between the straight parallel lines on a hilt, and the standard deviation of that average.
X-Y. WaveH, StDev_WaveH: The average height of the wave decorations on a hilt, and the standard deviation of that average.
Z-AA. WaveW, StDev_WaveW: The average width of the wave decorations on a hilt, and the standard deviation of that average.
AB-AC. WaveDist, StDev_WaveDist: The average distance between the wave decorations on a hilt, and the standard deviation of that average.
AD-AE. DashL, StDev_DashL: The average length of dashes on a hilt, and the standard deviation of that average.
AF-AG. DashDist, StDev_DashD: The average distance between dashes on a hilt, and the standard deviation of that average.
AH-AJ. BPrin1-3: The first three principle components of the blade curvature, calculated using the PrinComp program provided with SHAPE V1.3. These describe 95% of the variation between the profiles.
AK-AO. CSPrin1-5: The first five principle components of the cross section curvature, calculated using the PrinComp program provided with SHAPE V1.3. These describe 95% of the variation between the profiles.
AP-DQ. HTP A1 - D20: All of sine and cosine coefficients that result in the Fourier transform for the hilt top profile, as output by CHC2NEF in SHAPE V1.3.
DR-GS. HSP A1 - D20: All of sine and cosine coefficients that result in the Fourier transform for the hilt side profile, as output by CHC2NEF in SHAPE V1.3.