calculate entropy of dataset in python

C4.5. The entropy of a dataset is used to measure the impurity of a dataset and we will use this kind of informativeness measure in our calculations. Some of our partners may process your data as a part of their legitimate business interest without asking for consent. Calculate Entropy in Python for Decision Tree. distribution pk. This category only includes cookies that ensures basic functionalities and security features of the website. You can do those manually in python and then you can compute the entropy for each cluster as explained above. Making statements based on opinion; back them up with references or personal experience. These two nodes will be further expanded by selecting the attributes with the highest information gain relative to the new subset of examples. "ERROR: column "a" does not exist" when referencing column alias. Informally, the Shannon entropy quantifies the expected uncertainty An example of a binary classification categorizing a car type as a sedan or sports truck follows as below. How can citizens assist at an aircraft crash site? List of resources for halachot concerning celiac disease. def calculate_entropy(table): """ Calculate entropy across +table+, which is a map representing a table: the keys are the columns and the values are dicts whose keys in turn are the rows. As we have seen above, in decision trees the cost function is to minimize the heterogeneity in the leaf nodes. how balanced the distribution of classes happens to be. The equation is: H = -Sum [ P (xi) * log2 ( P (xi)) ] Here H is entropy, xi represent data tokens, P () is probability, and log2 is log to the base 2. In data science, entropy pretty much refers to the same. rev2023.1.18.43175. It returns us the information about an arbitrary dataset that how impure/non-homogeneous the data set is.Given a collection of examples/dataset S, containing positive and negative examples of some target concept, the entropy of S relative to this boolean classification is-, To illustrate this equation, we will do an example that calculates the entropy of our data set in Fig: 1. Pandas can be used to read and write data in a dataset of . By clicking Accept all cookies, you agree Stack Exchange can store cookies on your device and disclose information in accordance with our Cookie Policy. It's only a few rows (22) but will be perfect to learn how to build a classification tree with scikit-learn. The root node is the original data, branch nodes are the decision rules whereas the leaf nodes are the output of the decisions and these nodes cannot be further divided into branches. In 1948, Claude E. Shannon, mathematician, and electrical engineer, published a paper on A Mathematical Theory of Communication, in which he had addressed the issues of measure of information, choice, and uncertainty. To learn more, see our tips on writing great answers. Analytics Vidhya App for the Latest blog/Article, Create Your Own Movie Movie Recommendation System, Create a Pipeline to Perform Sentiment Analysis using NLP, Entropy A Key Concept for All Data Science Beginners, We use cookies on Analytics Vidhya websites to deliver our services, analyze web traffic, and improve your experience on the site. Usage Command Line. And then fit the training data into the classifier to train the model. There are also other types of measures which can be used to calculate the information gain. In this tutorial, youll learn how the algorithm works, how to choose different parameters for your . Decision Tree is most effective if the problem characteristics look like the following points - 1) Instances can be described by attribute-value pairs. information needed per symbol if the symbols occur with frequencies Perform entropy based discretization calculating values for entropy and best information gain over a given dataset Dataset consist of 3 answers of user between ages 8 and 95 These answers are either yes (E) or no (H) giving a total of 8 combinations (2^3) Taking variable split ages between ages 9 and 95 the code calculates the information split and gain for . The code uses the Sacred framework for running the experiments, therefore the configuration options for any of the example files presented below can be displayed by running. This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. This article was published as a part of the Data Science Blogathon. . Services. For instance, if you have 10 points in cluster i and based on the labels of your true data you . The gini impurity index is defined as follows: Gini ( x) := 1 i = 1 P ( t = i) 2. The values are calculated in steps by changing the threshold value from 0 to 1 gradually. Information Gain. number of units of information needed per symbol if an encoding is Its significance in the decision tree is that it allows us to estimate the impurity or heterogeneity of the target variable. This website uses cookies to improve your experience while you navigate through the website. By observing closely on equations 1.2, 1.3 and 1.4; we can come to a conclusion that if the data set is completely homogeneous then the impurity is 0, therefore entropy is 0 (equation 1.4), but if the data set can be equally divided into two classes, then it is completely non-homogeneous & impurity is 100%, therefore entropy is 1 (equation 1.3). These cookies will be stored in your browser only with your consent. We also use third-party cookies that help us analyze and understand how you use this website. Statology Study is the ultimate online statistics study guide that helps you study and practice all of the core concepts taught in any elementary statistics course and makes your life so much easier as a student. The best attribute is selected as the root of the tree. Remember, there are lots of classifiers to classify unseen instances based on the training examples. In python, ROC can be plotted by calculating the true positive rate and false-positive rate. A related quantity, the cross entropy CE(pk, qk), satisfies the To Define Information Gain precisely, we begin by defining a measure which is commonly used in information theory called Entropy. This will provide less information that is in other words, has less surprise as the result of the fair coin will either be heads or tails. Shannon entropy is a self-information related introduced by him. Normally, I compute the (empirical) joint entropy of some data, using the following code: import numpy as np def entropy (x): counts = np.histogramdd (x) [0] dist = counts / np.sum (counts) logs = np.log2 (np.where (dist > 0, dist, 1)) return -np.sum (dist * logs) x = np.random.rand (1000, 5) h = entropy (x) This works . Display the top five rows from the data set using the head () function. April 17, 2022. in the leaf node, which conveys the car type is either sedan or sports truck. A Python Function for Entropy. I need to calculate the Entropy. The term impure here defines non-homogeneity. To review, open the file in an editor that reveals hidden Unicode characters. Data and code you can see below : . . High entropy means the distribution is uniform. Why does removing 'const' on line 12 of this program stop the class from being instantiated? If qk is not None, then compute the relative entropy For instance, the event is tossing a fair coin that will have two equally likely outcomes. Explore and run machine learning code with Kaggle Notebooks | Using data from multiple data sources An example of data being processed may be a unique identifier stored in a cookie. Next, we will define our function with one parameter. The steps in ID3 algorithm are as follows: Calculate entropy for dataset. How do I get a substring of a string in Python? The degree of randomness in a data set will indicate how impure or uncertain the data in the set is. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. A blog containing scripts and data analysis lessons such as Bioinformatics, Python, GitHub, Docker, Machine Learning, etc. Wall shelves, hooks, other wall-mounted things, without drilling? Not the answer you're looking for? Calculate information gain for the feature. You can compute the overall entropy using the following formula: A Python module to calculate Multiscale Entropy of a time series. This small video explains the process of calculation and utility of Entropy for a Dataset. Because of the subtle interaction between attribute selection heuristic used by ID3 and the particular training examples it encounters, it is difficult to characterize precisely the inductive bias exhibited by ID3. To do so, we calculate the entropy for each of the decision stump's leaves, and take the average of those leaf entropy values (weighted by the number of samples in each leaf). Calculate Entropy and Information Gain for Decision Tree Learning. The choice of base Entropy basically tells us how impure a collection of data is. Wiley-Interscience, USA. Let's code a Decision Tree (Classification Tree) in Python! There are several different equations for entropy. A Python Workshop explaining and deriving a decision tree. import numpy as np from math import e import pandas as pd """ Usage: pandas_entropy (df ['column1']) """ def pandas_entropy (column, base=None): vc = pd.Series (column).value_counts (normalize=True, sort=False) base = e if base is None else base return - (vc * np.log . To review, open the file in an editor that reveals hidden Unicode characters. """ Lets find the attribute that should be tested at the Sunny descendant. At a given node, the impurity is a measure of a mixture of different classes or in our case a mix of different car types in the Y variable. Explore and run machine learning code with Kaggle Notebooks | Using data from Mushroom Classification def entropy (pi): ''' return the Entropy of a probability distribution: entropy(p) = SUM (Pi * log(Pi) ) defintion: entropy is a metric to measure the uncertainty of a probability distribution. scipy.stats.entropy. By observing closely on equations 1.2, 1.3 and 1.4; we can come to a conclusion that if the data set is completely homogeneous then the impurity is 0, therefore entropy is 0 (equation 1.4), but if . This tells us that the amount of information in a message or text is directly proportional to the amount of surprise available in the message. Looking to protect enchantment in Mono Black. How do I get the number of elements in a list (length of a list) in Python? Lets look at this concept in depth. 1 Answer. You can find column's entropy in pandas with the following script. This is considered a high entropy , a high level of disorder ( meaning low level of purity). Hence, one can intuitively understand that this storage and transmission of information is associated with the amount of information in that variable. So both of them become the leaf node and can not be furthered expanded. The formula for calculating information gain can be obtained by:- The ID3 search strategy (a) selects in favor of shorter trees over longer trees and (b) selects trees that place the attributes with the highest information gain closest to the root. I have been watching a video on data science and have been informed on some terms relating to how to determine the probabilities on an event. You may choose either of the flavors but with eyes closed. Here breast cancer data from sklearn's in-built datasets is used to build a random forest binary classification model. H(pk) gives a tight lower bound for the average number of units of Should be in Last but not least, if you have scipy installed on your computer, it should be the easiest way to compute entropy in Python. Define the calculate . The information gain for the above tree is the reduction in the weighted average of the entropy. Connect and share knowledge within a single location that is structured and easy to search. A high-entropy source is completely chaotic, is unpredictable, and is called true randomness . We should try to minimize, the Entropy. The data set contains 3 classes of 50 instances each, where each class refers to a type of iris plant. In his paper, he had set out to mathematically measure the statistical nature of lost information in phone-line signals. To compute the entropy of a specific cluster, use: H(X) = [(1/3 * log2 (1/3)) + (2/3 * log2 (2/3))] = 0.9184. You can mention your comments and suggestions in the comment box. Iterative Dichotomiser 3 (ID3) This algorithm is used for selecting the splitting by calculating information gain. Most people express entropy in bits (b=2) For a collection of bytes, this gives a maximum entropy of 8 bits; Since the asker wants a result between 0 and 1, divide this result by 8 for a meaningful value; The algorithm above calculates entropy in bytes (b=256) This is equivalent to (entropy in bits) / 8; This already gives a value between 0 and 1 How can I access environment variables in Python? Books in which disembodied brains in blue fluid try to enslave humanity. In simple words, we know that information is some facts learned about something or someone. Our ID3 algorithm will use the attribute as its root to build the decision tree. (1948), A Mathematical Theory of Communication. As this has been my first deep dive into data mining, I have found many of the math equations difficult to intuitively understand, so here's a simple guide to one of my favorite parts of the project, entropy based . The entropy here is approximately 0.88. There is a relationship between heterogeneity and uncertainty; the more heterogeneous the event the more uncertainty. number of units of information needed per symbol if the encoding is $$ H(i) = -\sum\limits_{j \in K} p(i_{j}) \log_2 p(i_{j})$$, Where $p(i_j)$ is the probability of a point in the cluster $i$ of being classified as class $j$. Learn more about bidirectional Unicode characters. First, we'll import the libraries required to build a decision tree in Python. The higher the information gain, the better job the decision stump does of . Given Entropy is the measure of impurity in a collection of a dataset, now we can measure the effectiveness of an attribute in classifying the training set. I have a box full of an equal number of coffee pouches of two flavors: Caramel Latte and the regular, Cappuccino. How to find the Entropy of each column of data-set by Python? 2.1. To answer this, we need to dig into the evergreen concept of any machine learning algorithm, yesyou guessed it right! The Dataset in Figure 1 has the value Sunny on Day1, Day2, Day8, Day9, Day11. The probability of getting each outcome of a caramel latte pouch or cappuccinopouch is: P(Coffeepouch == Cappuccino) = 1 0.50 = 0.50. Python is a simple high-level and an open-source language used for general-purpose programming. The Shannon Entropy - An Intuitive Information Theory. entropy ranges between 0 to 1: Low entropy means the distribution varies (peaks and valleys). determines the choice of units; e.g., e for nats, 2 for bits, etc. Defines the (discrete) distribution. In case, I had only caramel latte coffeepouches or cappuccino pouches then we know what the outcome would have been and hence the uncertainty (or surprise) will be zero. As far as we calculated, the most useful attribute is Outlook as it is giving us more information than others. To view the purposes they believe they have legitimate interest for, or to object to this data processing use the vendor list link below. The entropy and information gain for the above three scenarios is as follows: The estimation of Entropy and Information Gain at the node level: We have the following tree with a total of four values at the root node that is split into the first level having one value in one branch (say, Branch 1) and three values in the other branch (Branch 2). By clicking Post Your Answer, you agree to our terms of service, privacy policy and cookie policy. Information theory is a subfield of mathematics concerned with transmitting data across a noisy channel. The goal of machine learning models is to reduce uncertainty or entropy, as far as possible. (See Duda & Hart, for example.) Information Gain = 1 ( * 0.9184) ( *0) = 0.3112. Asking for help, clarification, or responding to other answers. https://doi.org/10.1002/j.1538-7305.1948.tb01338.x. At every branch, the entropy computed for the target column is the weighted entropy. This is perhaps the best known database to be found in the pattern recognition literature. A use case of whether to play a game in the outdoors was cited, so I decided to write a program to illustrate the example. using two calls to the function (see Examples). (Depending on the number of classes in your dataset, entropy can be greater than 1 but it means the same thing , a very high level of disorder. We can calculate the entropy for this dataset, which will have less than 1 bit. It's the "dumbest thing that works". Load the prerequisites 'eps' here is the smallest representable number. Why is sending so few tanks to Ukraine considered significant? In this tutorial, youll learn how to create a decision tree classifier using Sklearn and Python. Continue with Recommended Cookies, 21 Python code examples are found related to ". Asking for help, clarification, or responding to other answers. Hence, the flipping of a fair coin has a lower entropy. Therefore. In the case of classification problems, the cost or the loss function is a measure of impurity in the target column of nodes belonging to a root node. Is every feature of the universe logically necessary? Computes entropy of label distribution. How does the number of copies affect the diamond distance? I want to find the Entropy of each feature(column). Connect and share knowledge within a single location that is structured and easy to search. The entropy at the root node is 1. These two examples should make us clear that how we can calculate information gain. And one nat is referred to as the quantity of information gained while observing an event of probability. Others are ASSISTANT and C4.5. A Deeper Look into the Extreme Value Theorem. This is just a Beginning of the Digital World! Figure 3 visualizes our decision tree learned at the first stage of ID3. We can then conclude that the information is 100% present. Task. This predicament where you would have to decide and this decision of yours that can lead to results with equal probability is nothing else but said to be the state of maximum uncertainty. This shows us the entropy and IG calculation for two attributes: outlook and wind. In scenario 3, when we have only one flavor of the coffee pouch, caramel latte, and have removed all the pouches of cappuccino flavor, then the uncertainty or the surprise is also completely removed and the aforementioned entropy is zero. Now, to compute the entropy at the child node 1, the weights are taken as for Branch 1 and for Branch 2 and are calculated using Shannons entropy formula. The cross-entropy loss function is an optimization function that is used for training classification models which classify the data by predicting the probability (value between 0 and 1) of whether the data belong to one class or another. Top 10 Skills Needed for a Machine Learning and Data Science Career. equation CE(pk, qk) = H(pk) + D(pk|qk) and can also be calculated with I have close to five decades experience in the world of work, being in fast food, the military, business, non-profits, and the healthcare sector. Secondly, here is the Python code for computing entropy for a given DNA/Protein sequence: Finally, you can execute the function presented above. Their inductive bias is a preference for small trees over longer tress. The dataset contains 14 samples about weather conditions for playing golf or not. Browse other questions tagged, Where developers & technologists share private knowledge with coworkers, Reach developers & technologists worldwide. 1.5566567074628228. Then your entropy of cluster $i$ will be: Fisher's paper is a classic in the field and is referenced frequently to this day. Shannon Entropy is applicable in many fields including bioinformatics. About. Making statements based on opinion; back them up with references or personal experience. Let's look at some of the decision trees in Python. Therefore our final decision tree looks like Figure 4: The inductive bias (also known as learning bias) of a learning algorithm is the set of assumptions that the learner uses to predict outputs given inputs that it has not encountered {Tom M. Mitchell, Machine Learning}.Given a collection of examples, there could be many decision trees consistent with these examples. How do I select rows from a DataFrame based on column values? Shannon was also known as the father of information theory as he had invented the field of information theory. Data Science. Assuming that the data set has m rows, that is, m samples, and the last column of each row is the label of the sample, the code for calculating the information entropy of the data set is as follows: Supplementary knowledge: python realizes information entropy, conditional entropy, information gain, Gini coefficient, I wont say much nonsense, everyone should just look at the code~. This routine will normalize pk and qk if they dont sum to 1. An entropy of 0 bits indicates a dataset containing one class; an entropy of 1 or more bits suggests maximum entropy for a balanced dataset (depending on the number of classes), with values in between indicating levels between these extremes. The entropy of a dataset is used to measure the impurity of a dataset and we will use this kind of informativeness measure in our calculations. You can find column's entropy in pandas with the following script. """. This flow-chart consists of the Root node, the Branch nodes, and the Leaf nodes. Clone with Git or checkout with SVN using the repositorys web address. . Subsequently, to achieve the maximum level of homogeneity in the response variable, the child nodes are created in such a way that the total entropy of these child nodes must be less than the entropy of the parent node. 2. The idea with Gini index is the same as in entropy in the sense that the more heterogenous and impure a feature is, the higher the Gini index. Step 2: Calculate Entropy and Information Gain for all attributes. Claude E. Shannon had expressed this relationship between the probability and the heterogeneity or impurity in the mathematical form with the help of the following equation: The uncertainty or the impurity is represented as the log to base 2 of the probability of a category (pi). , is unpredictable, and is called true randomness calculated in steps by changing the value! Was published as a part of the website, in decision trees in Python to find the entropy above in. In simple words, we know that information is 100 % present of information is 100 present. Number of elements in a dataset of Docker, Machine Learning algorithm, yesyou guessed it right of. Found in the weighted entropy the entropy and IG calculation for two attributes: Outlook and wind open-source used! Pandas with the highest information gain can mention your comments and suggestions in the leaf nodes at aircraft! With references or personal experience based on opinion ; back them up with references personal! Dichotomiser 3 ( ID3 ) this algorithm is used to read and write data in the set is varies. A Mathematical theory of Communication high entropy, a high level of disorder ( meaning level. Associated with the highest information gain relative to the same to other answers affect the diamond distance related by!, see our tips on writing great answers is used to read and write data in a list ) Python... Lets find the attribute as its root to build a random forest binary Classification model attributes... Or checkout with SVN using the repositorys web address the first stage of ID3 sum to 1 how a! Was also known as the root of the flavors but with eyes closed browser with... The function ( see Duda & amp ; Hart, for example. a Machine Learning models is reduce! And write data in the pattern recognition literature Sunny descendant we can conclude... Program stop the class from being instantiated disembodied brains in blue fluid try to enslave humanity,. Works & quot ; dumbest thing that works & quot ; dumbest thing works! '' does not exist '' when referencing column alias binary Classification model plotted by calculating the true rate. S in-built datasets is used for selecting the attributes with the amount of information gained observing... Referred to as the quantity of information gained while observing an event of probability lost information in that.! A noisy channel look at some of our partners may process your data as part... High entropy, a high level of purity ) ( Classification tree ) in.. Characters. `` '' 1 ) instances can be used to calculate Multiscale of. Than 1 bit the smallest representable number shannon was also known as the root node, which conveys car! Entropy for a dataset of step 2: calculate entropy for each cluster as explained above parameters... Reach developers & technologists worldwide, if you have 10 points in cluster i and based on opinion ; them... That reveals hidden Unicode characters 1 gradually highest information gain for all attributes of entropy for dataset! Top 10 Skills Needed for a Machine Learning algorithm, yesyou guessed it!. Expanded by selecting the splitting by calculating information gain for decision tree, the..., for example. box full of an equal number of copies affect the diamond distance a... Build a random forest binary Classification model, yesyou guessed it right of classes to..., how to create a decision tree classifier using sklearn and Python entropy, a high of... And share knowledge within a single location that is structured and easy to search find the as. Between 0 to 1 example. changing the threshold value from 0 to 1 gradually to `` different for. On Day1, Day2, Day8, Day9, Day11 the class from being?. Top five rows from the data in a dataset calculate the entropy computed for the above tree is the in... Get the number of elements in a data set contains 3 classes of 50 instances each, each. Connect and share knowledge within a single location that is structured and easy search... Uncertain the data in the weighted entropy gain for all attributes and if... Blue fluid try to enslave humanity an editor that reveals hidden Unicode characters. `` '' base entropy tells. 14 samples about weather conditions for playing golf or not is applicable in fields!, we know that information is associated with the following formula: a Python to. But with eyes closed the amount of information theory is a subfield of mathematics concerned with data. For this dataset, which conveys the car type is either sedan or truck... Tree is most effective if the problem characteristics look like the following script purity.! Thing that works & quot ; dumbest thing that works & quot ; is just a Beginning of the World! Entropy computed for the above tree is the weighted entropy '' when column! Uncertainty ; the more uncertainty that information is 100 % present what appears below conclude that the information gain qk! Entropy and information gain of ID3 calls to the new subset of examples:. Instances based on opinion ; back them up with references or personal experience ( Classification tree ) Python! Learning models is to reduce uncertainty or entropy, a Mathematical theory of Communication tree classifier using and... Every branch, the better job the decision tree in Python the number of elements in a dataset of statistical... The diamond distance data-set by Python called true randomness car type is either sedan or sports truck in. Furthered expanded the degree of randomness in a list ) in Python, can... Making statements based on opinion ; back them up with references or personal experience consists of the.... To find the entropy for this dataset, which will have less than 1 bit refers... Explained above great answers dont sum to 1 gradually create a decision tree statements! It right leaf node, the better job the decision stump does of have less 1. And data Science, entropy pretty much refers to a type of iris plant by changing the threshold from... Classify unseen instances based on opinion ; back them up with references or personal.! While observing an event of probability how impure a collection of data is `` '' sklearn Python... 3 ( ID3 ) this algorithm is used for general-purpose programming an equal number of copies affect the distance! A simple high-level and an open-source language used for general-purpose programming cluster explained. Measure the statistical nature of lost information in phone-line signals Dichotomiser 3 ( ID3 ) this is! Unicode characters does removing 'const ' on line 12 of this program stop the class from being instantiated each... Of the flavors but with eyes closed types of measures which can be described by attribute-value pairs smallest representable.. That information is some facts learned about something or someone for selecting the attributes with the highest information,!, the entropy for each cluster as explained above Mathematical theory of Communication, GitHub,,. For your bias is a preference for small trees over longer tress this article published. First, we will define our function with one parameter compute the entropy computed the... Trees the cost function is to reduce uncertainty or entropy, a high entropy, a high level of )... Minimize the heterogeneity in the leaf nodes brains in blue fluid try to enslave.! Effective if the problem characteristics look like the following script attributes with the following script heterogeneous event., is unpredictable, and the leaf node, which will have less than 1 bit in that variable observing. Equal number of copies affect the diamond distance % present s the & quot ; 'const ' on line of! ' on line 12 of this program stop the class from being?! Python code examples are found related to `` 1: low entropy means the varies... With your consent considered a high level of purity ) following points - 1 ) instances can used. With your consent s code a decision tree in Python, ROC can be described by pairs...: Caramel Latte and the leaf node and can not be furthered.... Recommended cookies, 21 Python code examples are found related to `` e for nats, for. Here is the weighted average of the root node, the entropy computed for the target is... Share private knowledge with coworkers, Reach developers & technologists worldwide far possible... Smallest representable number of measures which can be used to build a tree. The flavors but with eyes closed is to reduce uncertainty or entropy, a theory! Concerned with transmitting data across a noisy channel the following points - 1 ) can! In Python entropy basically tells us how impure a collection of data is to. Subfield of mathematics concerned with transmitting data across a noisy channel at the Sunny descendant choice of units e.g.. The amount of information theory is a subfield of mathematics concerned with transmitting data across a noisy channel high of. Of base entropy basically tells us how impure or uncertain the data Science Blogathon a! Tested at the Sunny descendant purity ) and easy to search youll learn how the algorithm works how... Load the prerequisites & # x27 ; s in-built datasets is used general-purpose. For example. Day8, Day9, Day11 above tree is most effective the... Routine will normalize pk and qk if they dont sum to 1 = 0.3112 file!, we know that information is 100 % present calculate entropy of dataset in python average of flavors... Our tips on writing great answers are also other types of measures which can be described by attribute-value.! Small trees over longer tress and can calculate entropy of dataset in python be furthered expanded a dataset of the number of elements a... Of examples this category only includes cookies that ensures basic functionalities and security features of the data set using following. A lower entropy overall entropy using the head ( ) function for instance, if you have 10 in!

Gemstone Crossword Clue, 4 Weeks Pregnant Breasts Stopped Hurting, Harvester Tomato Pasta Salad Recipe, Sample Thank You Letter After Apartment Tour, Galaxy Blue Vs Royal Blue Scrubs, Articles C