DataScience Workbook / 09. Project Management / 4. Quality Assurance / 4.3 Research Reproducibility
Introduction
At its core, reproducibility is the ability to repeat a research study’s processes and obtain the same results. It’s a hallmark of high-quality scientific work, ensuring that findings are not just a one-time miracle but a reliable outcome that others can achieve independently. This involves everything from the initial hypothesis to the methodologies, data analysis, and even the presentation of results. Decent documentation plays a crucial role in this process, serving as the foundation that enables others to understand, evaluate, and replicate the research findings accurately.
Replicability refers to the ability to achieve similar outcomes by conducting the same study but with different methods, data sets, or conditions. While it sounds similar, replicability extends the idea of reproducibility to test the universality of findings across various contexts.
Transparency is the openness with which researchers share their methods, data, and findings, enabling others to scrutinize and understand their work fully.
Why does reproducibility matter so much?
Reproducibility is not just a technical requirement. It’s a commitment to scientific excellence and integrity.
- Reproducibility is the bedrock upon which the trustworthiness of research is built.
- When studies are reproducible, they contribute to a body of knowledge that is reliable and actionable.
- It also ensures that research funds and efforts are utilized effectively, paving the way for standardized approaches and genuine advancements in science.
- This fosters further investigation, innovation, and application of research findings in real-world scenarios, from healthcare to environmental policy.
Practically, reproducibility enhances the credibility and impact of your research, establishing you as a trustworthy figure in your field. For your future research, it opens doors to collaboration and increases the likelihood of funding. It also allows you to efficiently revisit and extend your own work in the future, whether it's applying the methodology to new datasets or refining your analysis as new information becomes available.
Common Challenges
Despite its importance, achieving reproducibility is not without its hurdles. Some of the common challenges include:
-
Data Accessibility
Limited access to the original data can impede the ability of other researchers to reproduce results. -
Methodological Transparency
Without detailed documentation of research methodologies, reproducing studies becomes a game of guesswork. -
Software and Tools
The use of proprietary software or specialized tools that are not widely available can restrict reproducibility. -
Publication Bias
The tendency to publish positive results over negative or inconclusive ones can skew the understanding of a research area’s reproducibility.
From a researcher’s perspective, meeting reproducibility standards is challenging due to the extra time and effort required for detailed documentation, method sharing, and navigating data privacy concerns. The rush to publish quickly and the competitive academic landscape can push these tasks lower on the priority list. Without clear rewards or direct incentives for these efforts, motivation to invest in reproducibility can wane, despite its importance for scientific integrity.
Lessons Learned from Real-Life Examples
The path toward reproducible research is illuminated by the lessons gleaned from both its triumphs and tribulations. From groundbreaking collaborations to eye-opening challenges, the real-life examples offer invaluable lessons on the critical importance of transparency, meticulous documentation, and community standards in ensuring research integrity and impact.
Let’s explore a few research stories to distill the wisdom gained from their experiences.
| Successful Examples of Reproducible Research | Challenges Due to Lack of Reproducibility |
|---|---|
| The Polymath Project [source ⤴] An open collaboration among mathematicians to solve complex problems, the Polymath Project exemplifies reproducibility in action. By documenting every hypothesis, computation, and discussion in real time on a public blog and wiki, the project has not only solved significant problems but also set a standard for openness and reproducibility in mathematical research. see also Massively collaborative mathematics, Nature volume 461, pages 879–881 (2009) |
Psychology [source ⤴] Spearheaded by the Center for Open Science, this project aimed to replicate 100 psychological studies. The findings, published in 2015, showed that less than half (~39%) of the original findings were replicated based on predefined criteria. This initiative revealed the complexity of reproducibility issues in psychological, prompting a widespread reevaluation of research practices in the field. see also Over half of psychology studies fail reproducibility test, Nature (2015) |
| Human Microbiome Project (HMP) [source ⤴] This project, aimed at understanding the microbial components of the human genetic and metabolic landscape, has been a model of reproducibility. By making all data and methodologies publicly available through the HMP DACC (Data Analysis and Coordination Center), researchers worldwide have been able to validate findings and further investigate the human microbiome, accelerating scientific discovery. see also The Integrative Human Microbiome Project, Nature volume 569, pages 641–648 (2019) |
Cancer Research [source ⤴] A notable study by researchers at Amgen attempted to replicate 53 landmark studies in cancer biology, only to find a reproducibility rate of about 6%. This stark revelation shed light on the reproducibility challenges within biomedical research, emphasizing the need for more stringent documentation, sharing practices and validation of experimental methods and findings. This initiative prompts calls for improved standards and transparency in research. see also Half of top cancer studies fail high-profile reproducibility effort, Nature (2021) |
Recent publications continue to highlight the challenge of reproducibility in research, underscoring the ongoing need for a shift in our [reserchers] approach to embrace more transparent and rigorous practices.
Lessons Learned
The lessons learned from both successes in reproducibility and the challenges faced due to its absence highlight the essential practices and mindset shifts needed to advance the reliability and validity of scientific findings.
-
Importance of Open Data and Methodologies
The successes and failures in reproducibility underscore the importance of open access to data and detailed documentation of methodologies. Projects like HMP demonstrate how shared data can catalyze progress, while cases like the Reproducibility Project in Psychology reveal the pitfalls of insufficient transparency. -
Need for Robust Statistical Analysis
Many reproducibility issues arise from statistical misinterpretation or the use of inappropriate analysis methods. A critical lesson is the need for rigorous statistical scrutiny and the adoption of practices like pre-registration of studies to ensure clarity and integrity in research design. -
Collaboration and Community Standards
Both successful and challenging examples highlight the role of community engagement and the establishment of standards in fostering reproducibility. Collaborative efforts, such as the Polymath Project, illustrate how collective wisdom and open dialogue can lead to breakthroughs, while initiatives like the Reproducibility Project help set benchmarks for research practices.
By reflecting on these real-life examples, you can better navigate the complexities of reproducibility, adopting practices that enhance the reliability, integrity, and impact of your work.
Key Principles and Best Practices
In a research project, reproducibility is not just a goal but a guiding principle. It ensures that research can be trusted, built upon, and extended by the scientific community. By embedding transparency, accessibility, and adherence to standards into our research practices, we not only enhance the reproducibility of our work but also its impact and credibility. Adopting these principles requires a shift in mindset: a commitment to conducting research that is as open and accessible as well as systematic and comprehensively documented.
Transparency in Methodology
The first cornerstone of reproducible research is transparency in methodology. This means providing a clear, detailed account of how research is conducted, from the initial hypothesis through to data collection and analysis. Every step should be documented so thoroughly that fellow researchers can follow the same path without ambiguity. This includes specifying the software and tools used, the versions of those tools, and any custom code or algorithms developed.
This table provides clear examples of practices that promote transparency in the methodology versus those that hinder it, helping to guide you toward more reproducible research practices.
| Methodology | Good Practices | Bad Practices | Recommendations |
|---|---|---|---|
| Documentation | Creating comprehensive, step-by-step documentation of the research process, including hypothesis formulation, experimental design, and data analysis, in a publicly accessible format. | Providing vague descriptions of the research process or omitting key steps such as data preprocessing or statistical analysis techniques. | Adopt a standardized template for research documentation that includes sections for all critical stages of the research process. Use open platforms like GitHub or OSF to share documentation. |
| Data Collection & Analysis | Sharing raw data sets along with detailed explanations of data collection methods, including instruments used, settings, and calibration information. | Failing to disclose specific data collection parameters or modifying data without transparently documenting changes and rationales. | Utilize data repositories with metadata standards to share raw data and detailed collection methodologies. Document any data transformations or cleaning steps in a reproducible manner. |
| Software & Tools | Publishing the exact versions of software and libraries used, along with custom scripts or code in open repositories, including necessary comments for understanding. | Using proprietary or custom software without providing access or sufficient detail for others to recreate the computational environment. | Whenever possible, opt for open-source alternatives to proprietary software. Clearly document the setup of your computational environment, including software versions and settings. |
| Methodological Choices | Explaining the rationale behind methodological choices, including the selection of control variables, statistical models, and the handling of outliers. | Making arbitrary methodological decisions without explanation or ignoring established protocols without justification. | Provide a clear justification for each methodological choice, referencing existing literature or preliminary results. If deviating from standard protocols, explain why this approach better suits the research objectives. |
EXERCISE
This exercise is designed to enhance your awareness and implementation of transparency in research methodologies, fostering a culture of openness and replicability.
Objective:
Create a customized checklist that ensures transparency in the methodology of your projects.
Instructions:
1. List all steps of your typical research process.
2. Next to each step, jot down specific questions that ensure transparency.</br> 3. Use this checklist as a guide when documenting your methodology in future projects.
See General Example
Study Design
Participant Selection
Data Collection Procedures
Statistical Analysis
Replicability Check
Ethical Considerations
See Specific Example
Project Context: A bioinformatics study aiming to identify novel drug targets in breast cancer by analyzing differential gene expression between tumor and normal tissue samples using RNA-Seq data.
Hypothesis Formulation
Have I clearly stated the biological question?
i.e., Identifying potential drug targets by analyzing differential gene expression in breast cancer
Experimental Design
Is the selection criteria for tumor vs. normal tissue samples fully described, including patient demographics and clinical criteria?
Data Collection
Have I detailed:
Data Analysis
Preprocessing:
Statistical Methods
Is the approach for adjusting p-values for multiple comparisons (using Benjamini-Hochberg correction) and defining significance thresholds described?
Replicability Check
Does the documentation cover my customized visualization scripts in R and various mid-steps in Bash?
Ethical Considerations
Accessibility of Data and Materials
Accessibility is the second key principle, focusing on the ease with which data and materials can be obtained and used by others. This involves not only making your data available but ensuring it is in a format that is understandable and usable. Sharing should be done through reputable, accessible platforms or repositories that ensure data preservation and accessibility over time. This principle also encompasses the sharing of materials used in the research, such as chemical compounds, biological specimens, or software code.
This table aims to guide you toward best practices in making your data and materials accessible, while also highlighting common pitfalls. It also offers actionable advice to improve accessibility and long-term preservation to ensure that data remains usable for future research endeavors.
| Category | Good Practices | Bad Practices | Recommendations |
|---|---|---|---|
| Data Sharing | Making datasets publicly available in accessible, well-documented formats on platforms like Figshare or Dryad. | Keeping datasets private or sharing them in formats that are difficult to access or use. | Choose accessible, reputable data repositories that support long-term data preservation and provide clear documentation for datasets. |
| Material Sharing | Providing detailed protocols for experiments, along with information on how to access any physical or digital materials used. | Withholding experimental materials or not providing detailed information on how they can be obtained. | Utilize repositories for sharing biological materials, chemical compounds, or software code, ensuring all materials are accompanied by detailed usage protocols. |
| License Selection | Choosing open licenses that facilitate the reuse and reproduction of data and materials, such as Creative Commons for data and MIT or GPL for software. | Applying restrictive licenses that limit the use, modification, or sharing of data and materials. | Review and select licenses that support open science, clearly communicate these choices, and provide guidance on how to attribute and reuse materials. |
| Documentation Quality | Supplying comprehensive metadata and documentation for datasets and materials, including collection methods, conditions, and any preprocessing steps. | Providing inadequate or no metadata, making it difficult for others to understand or use the shared resources effectively. | Adopt standards for metadata and documentation within your field, and use tools that facilitate the creation of detailed, structured metadata. |
| Data Preservation & Accessibility Over Time | Ensuring long-term preservation and accessibility of data by using sustainable, stable platforms and adhering to best practices in digital preservation. | Storing data in formats or locations that are prone to obsolescence, risking data loss or inaccessibility over time. | Invest in durable, future-proof storage solutions and follow digital preservation guidelines to maintain data usability for the long haul. Regularly review and update storage practices to keep pace with technological advances. |
EXERCISE
This example is designed to help you systematically think through and plan for the accessibility of your research data and materials, ensuring they can be shared and used by others in the scientific community.
Objective:
Draft a data management plan (DMP) that outlines how data and materials will be handled, stored, and shared. Following this plan enhances the accessibility of your data and materials.
Instructions:
1. Identify the types of data you typically generate and list them.
2. For each data type, write down how you will store, archive, and share this data, considering data formats, platforms for sharing, and any privacy concerns.
3. Include a section on the tools and software you'll use, ensuring they align with open standards.
See General Example
Data Formats
Repositories
Storage and Backup
Sharing and Licensing
Privacy and Ethics
See Specific Example
Data Types & Formats
Data Storage and Archiving
Sharing and Licensing
Tools and Software
Documentation of all bioinformatics tools (e.g., STAR v2.7.3a, DESeq2 v1.28.1) used, including parameters, will be made available on GitHub and linked in the GEO submission.
Standards and Protocols
The third principle, adherence to standards and protocols, serves as a shared language in the scientific community, enabling researchers to communicate their methods and findings more effectively. By following established guidelines and practices, researchers can create a common framework that provides a clear and accessible blueprint of the study. It facilitates the replication of their work by others and ensures that outcomes are verifiable. This includes following established protocols and standards within your field. It’s also about contributing to the community by potentially setting new standards that others can follow, enhancing the collective ability to achieve reproducible results. By utilizing open-source tools and platforms, you ensure that your methods are not only transparent but accessible to anyone with an interest in your work. This not only enhances the credibility and validity of the research but also fosters collaboration and innovation.
This table highlights the importance of adhering to standards and developing robust protocols to enhance the reproducibility of research.
| Category | Good Practices | Bad Practices | Recommendations |
|---|---|---|---|
| Adherence to Standards | Following established international or discipline-specific standards for data formatting, analysis, and reporting. | Ignoring existing standards or inconsistently applying them across different stages of research. | Identify and educate yourself on the relevant standards for your field; integrate their use from the outset of your research project. |
| Protocol Development | Developing clear, detailed protocols based on best practices, and sharing these publicly for validation and reuse. | Using ad-hoc or poorly documented protocols that are difficult for others to follow or replicate. | Invest time in protocol development and documentation; consider peer review of protocols and sharing via platforms like protocols.io. |
| Use of Open-Source Tools | Utilizing open-source software and tools that are freely available, ensuring no barriers to accessing the computational resources needed to replicate research. | Reliance on proprietary or custom-built software without providing access or alternatives for replication. | Whenever possible, choose open-source options for your research tools and clearly document any necessary configurations or modifications. |
| Interoperability | Ensuring data and research outputs are interoperable, allowing seamless integration and use across different systems and platforms. | Creating data or outputs in formats that are not widely supported, limiting their utility and accessibility. | Adopt widely supported data formats and standards that enhance interoperability; test your outputs across different platforms to ensure compatibility. |
| Reproducibility Checks | Regularly conducting reproducibility checks of your own research to ensure that your methodologies and results are robust and transparent. | Neglecting to verify the reproducibility of your own work, risking the publication of findings that cannot be independently validated. | Implement reproducibility checks as a standard part of your research process, including third-party validations where feasible. |
EXERCISE
This exercise is intended to encourage you to actively engage with and adhere to the standards and protocols in your field, ensuring that your research practices meet global benchmarks for quality and reproducibility.
Objective:
Evaluate and document the standards and protocols you currently use or plan to use in your research. In a result, you will get the assessment document that outlines your commitment to standards and protocols, ensuring reproducibility.
Instructions:
1. Make a list of the main standards and protocols relevant to your field of research (e.g., data formats, metadata standards).
2. Assess how well your current or planned research practices align with these standards.
3. For each standard or protocol, write a brief plan on how you will ensure compliance in your research.
See General Example
Current Practice Evaluation
Action Plan
Training Needs
See Specific Example
Data Formats
Use of FASTQ for raw data, BAM for alignments, adhering to the SAM/BAM format specifications.
Analysis Protocols
Following the ENCODE project's best practices for RNA-Seq analysis, including quality control steps and alignment guidelines.
Documentation Standards
Plan for Compliance
Comprehensive Documentation
Central to these principles is the need for comprehensive documentation. It’s not enough to follow these principles in practice; each component must be meticulously recorded and shared. It should cover the methodology, data sources, analysis techniques, and any challenges or deviations encountered along the way. Comprehensive documentation ensures that your research is not a closed book but an open invitation for scrutiny, collaboration, and advancement.
| Category | Good Practices | Bad Practices | Recommendations |
|---|---|---|---|
| Research Process Documentation | Meticulously documenting each step of the research process, from hypothesis formation to data analysis and interpretation, in a clear and accessible manner. | Offering only superficial descriptions of research steps or omitting details crucial for replication. | Adopt a structured documentation framework that covers all research stages and ensure it's easily accessible, possibly through open repositories. |
| Data Documentation | Providing comprehensive metadata for all datasets, including context, collection methods, and any preprocessing steps. | Leaving datasets without sufficient context or metadata, making them hard to interpret or reuse. | Use standardized metadata formats and repositories that encourage detailed data documentation to enhance usability and discovery. |
| Code and Software Documentation | Thoroughly commenting code and documenting software environments, including versions and dependencies, to facilitate replication. | Sharing code without comments or documentation, making it difficult for others to understand or execute. | Incorporate comprehensive commenting within your code and use tools like virtual environments to document software dependencies. |
| Methodological Justification | Clearly explaining the rationale behind methodological choices, highlighting how they contribute to the research objectives. | Making arbitrary methodological choices without transparent justification, confusing other researchers. | Provide a detailed rationale for each methodological decision, including references to supporting literature or preliminary results. |
| Challenges and Deviations | Proactively documenting challenges, obstacles, or deviations from the planned methodology, including how they were addressed or why certain decisions were made. | Ignoring or glossing over challenges and deviations, which can provide crucial context for understanding research outcomes. | Maintain a detailed log of challenges and deviations as part of your project documentation, emphasizing transparency and the reasoning behind key decisions. |
| Revision and Update Documentation | Documenting any revisions or updates to the research process, data, or analysis, along with the reasons for these changes. | Failing to update docs to reflect changes in the research process, leading to discrepancies between published work and actual practices. | Establish a version control system for your documentation to track changes over time and ensure that the latest information is always available. |
EXERCISE
This task is designed to guide you in creating a detailed and structured approach to documenting your research, making it easier for others to understand, replicate, and build upon your work.
Objective:
Develop a comprehensive documentation template tailored to your research process.
Instructions:
1. Outline the sections needed to fully document a project, including background, methodology, data collection and analysis, results, and deviations.
2. For each section, add prompts or questions that guide you to provide detailed information.
3. Integrate a section for notes on deviations or unexpected challenges encountered during the project.
See General Example
Methodology Detailed Description
Data Analysis
Results Summary
Deviations and Challenges
Conclusion and Future Directions
See Specific Example
Background
Detailed introduction to the research question, including the significance of identifying novel drug targets in breast cancer.
Data Collection and Analysis
A section dedicated to describing the data collection process, including:
Methodology
Results
Guidelines for presenting gene expression differences, including:
Challenges
Logging any deviations from the planned methodology, such as:
Tools and Technologies
Digital tools are invaluable for ensuring that research is transparent, accessible, and collaborative. They help create a seamless workflow from data collection to analysis, enabling researchers to document each step of their process in a manner that is clear and reproducible.
- By leveraging version control systems, researchers can track and manage changes over time, ensuring that every iteration of their work is preserved and accessible.
- Repositories for sharing datasets and code break down barriers to access, making it possible for scientists to build on each other’s work and collaborate in ways previously unimaginable.
- Open-source statistical and computational tools democratize the process of data analysis, allowing researchers from around the globe to use, modify, and improve upon existing methods.
Together, these digital tools form a critical infrastructure for advancing reproducible research, ensuring that findings are robust, verifiable, and ready to contribute to the collective knowledge of the scientific community.
Below, you can find tables summarizing essential tools for reproducible research, complete with quick links to practical tutorials to help you get started with each of them.
Version Control Systems
Version control systems ⤴ are essential for tracking changes in research projects, facilitating collaboration, and ensuring the integrity of research data and code over time.
| tool | tutorial | applications |
|---|---|---|
| Git | GIT - a distributed version control system GIT Commands |
Distributed version control system widely used for source code management. |
| SVN | Subversion (SVN) | Centralized version control system designed to keep track of changes in files and directories. |
Shared Repositories
Repositories for sharing datasets and code ⤴ are crucial for reproducibility, enabling researchers to access, use, and cite the exact datasets and software versions used in research studies.
| tool | tutorial | applications |
|---|---|---|
| GitHub | Introduction_to_GitHub GitHub for advanced users |
Hosting for software development and version control using Git, with features for collaboration, issue tracking, and code review. GitHub is the most commonly used platform for hosting and collaborating on open-source projects. |
| Zenodo | General-purpose open-access repository developed under the European OpenAIRE program, for research data and software. | |
| OSF | Open Science Framework supports the entire research lifecycle: planning, execution, reporting, archiving, and discovery. |
Open-source Tools
Open-source statistical and computational tools offer powerful, transparent, and community-validated methods for data analysis, contributing significantly to reproducible research practices.
| tool | tutorial | applications |
|---|---|---|
| R | Introduction to R programming and more ⤴ |
Statistical computing and graphics, with a wide range of packages for various data analysis tasks. |
| Python | Introduction to Python Programming and more ⤴ |
General-purpose programming language with extensive libraries for data analysis, machine learning, and scientific computing. |
| Julia | Introduction to Julia Programming and more ⤴ |
High-level, high-performance dynamic programming language for technical computing, with syntax that is familiar to users of other technical computing environments. |
| Jupyter Notebooks | Jupyter: Interactive Web-Based Multi-Kernel DE and more ⤴ |
Open-source web application that allows you to create and share documents that contain live code, equations, visualizations, and narrative text. |
Practice Your Skills
These exercises are crafted to provide you with practical skills and hands-on experience necessary for conducting reproducible research. By engaging with these tasks, you’ll not only learn to use essential tools but also understand how each contributes to the overall reproducibility of your research projects.
E1: Version Control Basics
This exercise is designed to illustrate the critical role of version control in tracking and managing changes in research projects, promoting collaboration, and ensuring that all project versions are accessible and reproducible.
Objective:
Familiarize yourself with Git for version control to manage changes in your research project efficiently.
Instructions:
1. Install Git: Download and install Git from https://git-scm.com/.
2. Create a Repository:
git init to initialize a new Git repository on your local machine.git add README.md to stage the file for committing.git commit -m "Initial commit".git log to view the history of changes.To see a more detailed step-by-step guide with graphical aids to complete this exercise, navigate to tutorial GIT - a distributed version control system ⤴ and follow the steps in the hands-on sections:
E2: Documenting with Jupyter Notebooks
This exercise is designed to underscore the importance of using open-source tools for data analysis, emphasizing clear, well-commented scripts that can be easily shared and replicated.
Objective:
Create a Jupyter Notebook that integrates code, outputs, and narrative, showcasing how this tool can enhance research documentation.
Instructions:
1. Install Jupyter: Follow the instructions at https://jupyter.org/install.
2. Create a New Notebook:
jupyter lab.To see a more detailed step-by-step guide with graphical aids to complete this exercise, navigate to tutorial Jupyter: Web-Based Programming Interface ⤴ and follow the installation steps in the Installing Jupyter ⤴ hands-on section. Then jump to Getting Started with JupyterLab ⤴ tutorial to launch Jupyter interface, learn about components in the GUI, create new notebook and add cells of various types. To start using rich text markup in your documentation, check out the Introduction to Markdown ⤴ tutorial.
Objective:
Perform a basic data analysis using R or Python, focusing on making the script reproducible.
Instructions:
1. Choose Your Tool
See Code Example
The Iris dataset is available in R by default through the datasets package, which is part of the standard R distribution. No additional installation is required for this exercise.
# Load the Iris dataset
data(iris)
# View the first few rows of the dataset
head(iris)
#2 Loading the Iris Dataset in Python
The Iris dataset can be loaded using Pandas. While Pandas is not in the Python Standard Library, it is widely used for data manipulation and analysis tasks. If you don't have Pandas installed, you can install it using the following command in the terminal:
pip install pandas seabornThen, you can load the the Iris dataset in a Jupyter Notebook or script file:
# Import pandas and seaborn
import pandas as pd
import seaborn as sns
# Load the iris dataset
iris = sns.load_dataset('iris')
# View the first few rows of the dataset
iris.head()
See Code Example (step 1)
Using R's built-in functions, we can quickly generate summary statistics and species counts, providing insights into the dataset's composition without the need for additional packages.
# Summary statistics
summary(iris)
# Count of species
table(iris$Species)
#2 Exploring the Iris Dataset in Python
Exploring the Iris dataset in Python requires the use of Pandas for data manipulation and analysis. Pandas' functions, such as
describe() for summary statistics and value_counts() for species counts, allow for an in-depth exploration of the dataset. (This step assumes Pandas is installed and imported.)# Summary statistics
iris.describe()
# Count of species
iris['species'].value_counts()
See Code Example (step 2)
#1 Ploting the Iris Dataset in R
To visualize the Iris dataset in R, we utilize the ggplot2 package, a powerful tool for creating complex plots from data in a dataframe. If ggplot2 is not already installed, it can be easily added.
# Install ggplot2 if not already installed
if(!require(ggplot2)) install.packages("ggplot2")
library(ggplot2)
# Scatter plot of Sepal.Length vs Sepal.Width colored by Species
ggplot(iris, aes(x = Sepal.Length, y = Sepal.Width, color = Species)) + geom_point() +
theme_minimal() + ggtitle("Sepal Length vs. Sepal Width")
#2 Ploting the Iris Dataset in Python
To plot the Iris dataset in Python, both matplotlib and seaborn libraries are essential. These libraries are not part of the Python Standard Library but can be installed via pip. Together, they provide a comprehensive toolkit for creating a variety of visualizations, including scatter plots. If needed, use the following command in your terminal:
pip install matplotlib seaborn# Import matplotlib for plotting
import matplotlib.pyplot as plt
import seaborn as sns # optional here, if loaded earlier along with pandas
# Scatter plot of Sepal.Length vs Sepal.Width colored by Species
sns.scatterplot(data=iris, x="sepal_length", y="sepal_width", hue="species")
plt.title("Sepal Length vs. Sepal Width")
plt.show()
To see a more detailed step-by-step guide with graphical aids to complete this exercise, explore the following tutorials:
Objective:
Explore collaborative features of GitHub, such as issues and pull requests, to understand community standards in code development.
Instructions:
1. Create a GitHub Account online: Sign up at https://github.com/ if you haven't already. (It is free.)
2. Ceate Your Own Empty Repository:
git remote add origin [repository-URL]
git push -u origin master or git push -u origin maindepending on your default branch name.
4. Share Your Notebook via GitHub:
Optional Steps for Further Engagement:
5. Fork a Repository: Explore other users' repositories related to your research interests, fork one, and consider contributing by adding improvements or additional analysis.
6. Engage with the Community: Participate in discussions, open issues for any bugs you find, or offer solutions to existing issues. Collaboration is key to advancing reproducible research.
To see a more detailed step-by-step guide with graphical aids to complete this exercise, navigate to tutorial GIT - a distributed version control system ⤴ and follow the steps in the Working with Remote Repos ⤴ hands-on section. You can engage further by exploring the Collaborating on Projects ⤴ section.