(paper) added NumPy citation per suggestion in #6

paper/paper.bib

@@ -296,3 +296,26 @@ @misc{Edmisten2022
    url = {https://wcedmisten.fyi/post/analyzing-all-recipes/#spices},
    year = {2022},
 }
+
+@Article{Harris2020array,
+ title         = {Array programming with {NumPy}},
+ author        = {Charles R. Harris and K. Jarrod Millman and St{\'{e}}fan J.
+                 van der Walt and Ralf Gommers and Pauli Virtanen and David
+                 Cournapeau and Eric Wieser and Julian Taylor and Sebastian
+                 Berg and Nathaniel J. Smith and Robert Kern and Matti Picus
+                 and Stephan Hoyer and Marten H. van Kerkwijk and Matthew
+                 Brett and Allan Haldane and Jaime Fern{\'{a}}ndez del
+                 R{\'{i}}o and Mark Wiebe and Pearu Peterson and Pierre
+                 G{\'{e}}rard-Marchant and Kevin Sheppard and Tyler Reddy and
+                 Warren Weckesser and Hameer Abbasi and Christoph Gohlke and
+                 Travis E. Oliphant},
+ year          = {2020},
+ month         = sep,
+ journal       = {Nature},
+ volume        = {585},
+ number        = {7825},
+ pages         = {357--362},
+ doi           = {10.1038/s41586-020-2649-2},
+ publisher     = {Springer Science and Business Media {LLC}},
+ url           = {https://doi.org/10.1038/s41586-020-2649-2}
+}

paper/paper.md

@@ -42,7 +42,7 @@ bibliography: paper.bib
 
 `nimCSO` is a high-performance tool implementing several methods for selecting components (data dimensions) in compositional datasets, which optimize the data availability and density for applications such as machine learning. Making said choice is a combinatorically hard problem for complex compositions existing in high-dimensional spaces due to the interdependency of components being present. Such spaces are encountered across many scientific disciplines (see [Statement of Need](#statement-of-need)), including materials science, where datasets on Compositionally Complex Materials (CCMs) often span 20-45 chemical elements, 5-10 processing types, and several temperature regimes, for up to 60 total data dimensions. This challenge also exists in everyday contexts, such as study of cooking ingredients [@Ahn2011], which interact in various recipes, giving rise to questions like *"Given 100 spices at the supermarket, which 20, 30, or 40 should I stock in my pantry to maximize the number of unique dishes I can spice according to recipe?"*. Critically, this is not as simple as frequency-based selection because, e.g., removing less common nutmeg and cinnamon from your shopping list will prevent many recipes with the frequent vanilla, but won't affect those using black pepper [@Edmisten2022].
 
-At its core, `nimCSO` leverages the metaprogramming ability of the Nim language [@Rumpf2023] to optimize itself at compile time, both in terms of speed and memory handling, to the specific problem statement and dataset at hand based on a human-readable configuration file. As demonstrated in the [Methods and Performance](#methods-and-performance) section, `nimCSO` reaches the physical limits of the hardware (L1 cache latency) and can outperform an efficient native Python implementation over 100 times in terms of speed and 50 times in terms of memory usage (*not* counting the interpreter), while also outperforming the NumPy implementation 37 and 17 times, respectively, when checking a candidate solution.
+At its core, `nimCSO` leverages the metaprogramming ability of the Nim language [@Rumpf2023] to optimize itself at compile time, both in terms of speed and memory handling, to the specific problem statement and dataset at hand based on a human-readable configuration file. As demonstrated in the [Methods and Performance](#methods-and-performance) section, `nimCSO` reaches the physical limits of the hardware (L1 cache latency) and can outperform an efficient native Python implementation over 100 times in terms of speed and 50 times in terms of memory usage (*not* counting the interpreter), while also outperforming the NumPy [@Harris2020array] based implementation 37 and 17 times, respectively, when checking a candidate solution.
 
 `nimCSO` is designed to be both (1) a user-ready tool, implementing two efficient brute-force approaches (for handling up to 25 dimensions), a custom search algorithm (for up to 40 dimensions), and a genetic algorithm (for any dimensionality), and (2) a scaffold for building even more elaborate methods in the future, including heuristics going beyond data availability. All configuration is done with a simple human-readable `YAML` file and plain text data files, making it easy to modify the search method and its parameters with no knowledge of programming and only basic command line skills.