mohamadkhalaj / datascience-cheatsheet

Cheatsheet for datascience libraries in python like Numpy, Pandas, Matplotlib and Scikit-learn.

Table of contents

Python

• Datatypes
• Difference between '==' and 'is'
• Some useful built-in functions
• List comprehension
• Lambda function
• Unpacking
• More about mutable and immutable datatypes

Numpy

• Why we use Numpy?
• Creating and inspecting array
• Indexing and slicing
• Operators
• Mask and filter
• Copy vs view
• Array manipulation
• Understanding axes
• Broadcasting
• Sparse matrix

Pandas

• What is Pandas?
• Series, dataframe and index
• Data indexing and selection
• Operators
• Doing some stuffs!
• Datatype
• Change datatype
• Describe
• Rename columns
• Rename rows
• Replace values
• Finding unique values
• Deleting column
• Deleting row
• Remove duplicates
• Mask and filter
• Apply funciton to the data
• Convert to numpy array
• Handling missing data
• Aggregation and grouping

Matplotlib

• Simple plot
• Title, Ticks, Labels, and Legends
• Colors, Markers, and Line Styles
• Subplot, Subplots and subplot2grid
• Histogram, bar, scatter, pie and heatmap
• Saving Plots to File

Scikit-learn

• What is machine learning?
• Preprocessing
• Identifying and handling the missing values
• Identifying outliers
• Encoding the categorical data
• Transforming the dataset
• Feature scaling
• Standardization
• Normalization
• Dimension reduction
• Principal Component Analysis (PCA)
• Train test split
• Feature selection
• Model selection
• Supervised learning
• Classification
• KNN
• SVM
• Random forest
• Decision tree
• Logistic regression
• Regression
• Linear regression
• Polynomial regression
• None linear regression
• Multiple linear regression
• Unsupervised learning
• Kmeans
• DBSCAN
• Semi supervised learning
• Model evaluation
• Classification
• Jaccard
• F1-score
• log-loss
• Clustering
• Sum of Squared Error (SSE) score
• Silhouette coefficient
• Prediction