Nikuradse (#175)

* Added two datasets from the Nikuradse 'Flow in Rought Pipes' paper.

* update dataset files

Created by https://github.com/heal-research/pmlb/actions/runs/4254088850\nfrom 5becab4 on 2023-02-23

* Update metadata.yaml files.

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

.lfs-assets-id

@@ -22,6 +22,7 @@
 0c342ef5d61bbcf43180a3b71d407b9d994942ce43e8960052201daf88dd095d
 0d05767a4c118752a25c4632aeea3b71ffa1bfe122b6a2401f85d20541be19a4
 0d39f17afc3a1712bd6c460aa941aed7835b3feb142538adfdd31ddc2451d60d
+0d43780ab866e54a2a78d8c86ba231ad0a5d55588450a33ed6fe52bee9638341
 0fa7cb761daaf9c9bac4ead348064cdb824db9ee3c6a7c968d73ae9a459469f6
 0fb9badfe474d0dbc5a23e83ed56dfc8cbd8ae6201f959f3ba0dd147fff89f09
 104e99890ea7e473e56d29c38452e79f173cbd5078cc30e1702eb45b42600585
@@ -314,6 +315,7 @@ b8d94f002408f850933b1539a7fa8530dcb68f0dccae47cd8658909e6dbe1146
 bbe46db284f0e29f95dfd9984ca32565b52d83217aa084c90de5f6edf82bfea1
 bd0e747cb0a16d9f68843ccd6fa0b0d382bb21f2c83ccbc222712426ee42274b
 bd9d5214451c3b72e8a5ba4ae75a565d373d90a804659d0a5f4617fad3ac4cfd
+be6942e13096c21f10496a24056b01ff791e24d6172f5b7a09013c3307d38f28
 beedc054b8e7d974a98326db8c834843eb188cffba0f07029a8370b193ce020e
 bf066d8b8431c89d3c8afd58b0bfe56f53e0aaedd8d4ec05c132268115af3f36
 bfc6131af9d009576a82d25e0590955980535eb61c67d1553434da993e79af92

datasets/nikuradse_1/README.md

@@ -0,0 +1,6 @@
+# nikuradse_1
+
+[**Pandas Profiling Report**](https://epistasislab.github.io/pmlb/profile/nikuradse_1.html)
+
+[Metadata](metadata.yaml) | [Summary Statistics](summary_stats.tsv)
+

datasets/nikuradse_1/SOURCE.txt

@@ -0,0 +1,30 @@
+The original source is "J. Nikuradse, Laws of Flow in Rough Pipes,
+Technical Memorandum 1292, National Advisory Committee for
+Aeronautics, 1950, (Translation of "Strömungsgesetze in rauhren
+Rohren", VDI-Forschungsheft 361, Beilage zu "Forschung auf dem Gebiete
+des Ingenierswesens" Ausgabe B Band 4, July/August 1933).
+Tables 2 to 8.
+
+
+The two input variables are
+ - relative roughness (column r/k)
+ - log of Reynolds number (column log Re).
+
+The target variable is log of 100 times resistance factor
+(column log(100 lambda)).  Figure 9 in the original Nikuradse paper
+shows additional points for Re < 10^3.8 but they are not contained in
+the tables.
+
+A separate version of this dataset has been used in [1].
+
+Another separate version of the data has been used in [2].
+
+
+[1] Bayesian Machine Scientist to Compare Data Collapses for the
+    Nikuradse Dataset Ignasi Reichardt, Jordi Pallarès, Marta
+    Sales-Pardo, and Roger Guimerà Phys. Rev. Lett. 124, 084503 -
+    Published 27 February 2020
+    https://doi.org/10.1103/PhysRevLett.124.084503
+
+[2] Bobby H. Yang & Daniel D. Joseph (2009) Virtual Nikuradse, Journal
+    of Turbulence, 10, N11, DOI: 10.1080/14685240902806491

datasets/nikuradse_1/metadata.yaml

@@ -0,0 +1,27 @@
+# Reviewed by Gabriel Kronberger
+dataset: nikuradse_1
+description: > 
+  Predict the resistance for fluid flow in rough pipes based on the Nikuradse dataset.
+source: > 
+  J. Nikuradse: Laws of Flow in Rough Pipes, Technical Memorandum 1292, National Advisory Committee for Aeronautics, 1950, (Translation of "Strömungsgesetze in rauhen Rohren", VDI-Forschungsheft 361, Beilage zu "Forschung auf dem Gebiete des Ingenierswesens" Ausgabe B Band 4, July/August 1933). Tables 2 to 8 and Figure 9.
+publication: >
+  Ignasi Reichardt, Jordi Pallarès, Marta Sales-Pardo, and Roger Guimerà: Bayesian Machine Scientist to Compare Data Collapses for the Nikuradse Dataset.  Phys. Rev. Lett. 124, 084503 - Published 27 February 2020 https://doi.org/10.1103/PhysRevLett.124.084503
+task: regression
+keywords:
+  - Physics
+  - Flow
+target:
+  type: continuous
+  description: log of 100x resistance factor (log(100 lambda))
+  code: null
+features:
+  - name: r_k
+    type: continuous
+    description: relative roughness (r/k), r is the radius of the pipe, k is the average projection of roughness.
+    code: null
+    transform: ~
+  - name: log_Re
+    type: continuous
+    description: log of Reynolds number (log Re).
+    code: null
+    transform: ~

datasets/nikuradse_1/summary_stats.tsv

@@ -0,0 +1,2 @@
+dataset	n_instances	n_features	n_binary_features	n_categorical_features	n_continuous_features	endpoint_type	n_classes	imbalance	task
+nikuradse_1	362	2	0	0	2	continuous	186.0	0.004936819240411923	regression

datasets/nikuradse_2/README.md

@@ -0,0 +1,6 @@
+# nikuradse_2
+
+[**Pandas Profiling Report**](https://epistasislab.github.io/pmlb/profile/nikuradse_2.html)
+
+[Metadata](metadata.yaml) | [Summary Statistics](summary_stats.tsv)
+

datasets/nikuradse_2/SOURCE.txt

@@ -0,0 +1,20 @@
+The original source is "J. Nikuradse, Laws of Flow in Rough Pipes,
+Technical Memorandum 1292, National Advisory Committee for
+Aeronautics, 1950, (Translation of "Strömungsgesetze in rauhren
+Rohren", VDI-Forschungsheft 361, Beilage zu "Forschung auf dem Gebiete
+des Ingenierswesens" Ausgabe B Band 4, July/August 1933).
+Tables 2 to 8.
+
+This data is for the Prandl collapse 1/sqrt(2) - 2 log(r/k)
+The input variable is log(v*k/nu). The target variable is 1/sqrt(2) - 2 log(r/k).
+
+Figure 11 in the original Nikuradse paper shows the data. The points
+in the figure do not match the data in the tables exactly.
+
+A separate version of this dataset has been used in [1].
+
+[1] Bayesian Machine Scientist to Compare Data Collapses for the
+    Nikuradse Dataset Ignasi Reichardt, Jordi Pallarès, Marta
+    Sales-Pardo, and Roger Guimerà Phys. Rev. Lett. 124, 084503 -
+    Published 27 February 2020
+    https://doi.org/10.1103/PhysRevLett.124.084503

datasets/nikuradse_2/metadata.yaml

@@ -0,0 +1,24 @@
+# Reviewed by Gabriel Kronberger
+dataset: nikuradse_2
+description: > 
+  Predict the resistance for fluid flow in rough pipes based on the Nikuradse dataset. This dataset uses the Prandl collapse.
+source: > 
+  J. Nikuradse: Laws of Flow in Rough Pipes, Technical Memorandum 1292, National Advisory Committee for Aeronautics, 1950, (Translation of "Strömungsgesetze in rauhen Rohren", VDI-Forschungsheft 361, Beilage zu "Forschung auf dem Gebiete des Ingenieurswesens" Ausgabe B Band 4, July/August 1933). Tables 2 to 8 and Figure 11.
+publication: > 
+  Ignasi Reichardt, Jordi Pallarès, Marta Sales-Pardo, and Roger Guimerà: Bayesian Machine Scientist to Compare Data Collapses for the Nikuradse Dataset.  Phys. Rev. Lett. 124, 084503 - Published 27 February 2020 https://doi.org/10.1103/PhysRevLett.124.084503
+task: regression
+keywords:
+  - Physics
+  - Flow
+target:
+  type: continuous
+  description: > 
+    Prandl collapse for the resistance factor (1/sqrt(2) - 2 log(r/k)), r is the radius of the pipe, k is the average projection of roughness.
+  code: null
+features:
+  - name: log_v_k_nu
+    type: continuous
+    description: > 
+      Input features for the Prandl collapse: log(v* k/nu), v* is the "friction" velocity, k is the average projection of roughness, nu is the kinematic velocity [cm^2 s^-1].
+    code: null
+    transform: ~

datasets/nikuradse_2/summary_stats.tsv

@@ -0,0 +1,2 @@
+dataset	n_instances	n_features	n_binary_features	n_categorical_features	n_continuous_features	endpoint_type	n_classes	imbalance	task
+nikuradse_2	362	1	0	0	1	continuous	80.0	0.020400530269280518	regression

pmlb/all_summary_stats.tsv

@@ -354,6 +354,8 @@ movement_libras	360	90	0	0	90	categorical	15.0	0.0	classification
 mushroom	8124	22	5	16	1	categorical	2.0	0.0012918910021209	classification
 mux6	128	6	6	0	0	categorical	2.0	0.0	classification
 new_thyroid	215	5	0	0	5	categorical	3.0	0.2990805840995132	classification
+nikuradse_1	362	2	0	0	2	continuous	186.0	0.0049368192404119	regression
+nikuradse_2	362	1	0	0	1	continuous	80.0	0.0204005302692805	regression
 nursery	12958	8	1	7	0	categorical	4.0	0.0900899366400157	classification
 optdigits	5620	64	3	10	51	categorical	10.0	1.4704868366802722e-05	classification
 page_blocks	5473	10	0	0	10	categorical	5.0	0.7627104291422278	classification