Master in Data Analysis-numpy, pandas, visualuze & Streamlit

Install Python from the official Python.org site, select the correct version for your operating system, and verify the installation using idle or Jupyter Notebook in this data analysis course.

Install the Jupyter notebook via the Anaconda distribution, selecting the distribution installer or Miniconda, and note the included libraries for computer vision and natural language processing.

Learn how to launch and navigate Jupyter notebook from Anaconda. Explore the local interface, toolbars, file system, and running code in notebook cells with a kernel.

Master Jupyter notebook workflow by opening and renaming notebooks, switching between code and markdown cells, and using shortcuts to run or create new cells.

Install and set up PyCharm community edition for Python development, create a new project, configure the environment, and run Python scripts to practice pip installs and terminal commands.

Explore data storage options such as NumPy, Pandas, and tensors, and learn why arrays, data frames, and tensors suit multi-dimensional and large datasets.

Discover what NumPy is, how to install it, and how to use it in a Jupyter notebook. Explore versions, import NumPy, and create arrays while comparing arrays and lists.

Explore numpy arrays properties, including homogeneous data types, array ordering, and n-dimensional structures; perform indexing, slicing, linear algebra, Fourier transforms, and random shuffling, while learning mutability and concatenation.

Learn to create numpy arrays in a Jupyter notebook by choosing lists or tuples, avoiding sets, and using numpy and Python aggregation functions like max, min, sum, and mean.

Compare Python lists and numpy arrays: lists are one-dimensional with mixed types, while numpy arrays support multi-dimensional data and enable arithmetic operations with memory efficiency for data science.

Explore the differences between vectors, matrices, and tensors—from one-dimensional feature vectors to multi-dimensional tensors, using age, experience, and salary data as examples, with AI frameworks TensorFlow and PyTorch.

Explore reshaping and resizing arrays from vectors to matrices and tensors by examining shapes like four by one and three by three, and compare reshape versus resize behavior.

Learn to create and initialize numpy arrays with zeros, ones, and full by specifying shape and dtype and optionally filling with custom values such as 22 by seven or 3.14.

Learn to create arrays with numpy arrange from a start to stop with a step, including evens with a step of 2, and reshape to 50x2, 25x4, or 10x10.

Learn to install and use pandas, verify versions, and access documentation. Build data frames and series, and load CSV or Excel files with read_csv.

Explore the differences between series and data frames in pandas, highlighting one-dimensional series versus two-dimensional data frames, labeling, and multi-type data storage.

Label columns and row indices in a pandas data frame by supplying custom column names and index labels when creating it from lists.

Load external data into a Jupyter notebook, use read_csv with pandas to create dataframes, and explore datasets like car prices.

Explore dataframes with the head and tail methods in pandas by loading a csv, inspecting top and bottom rows, and controlling the number of rows displayed.

Explore basic dataframe properties by inspecting column names, shapes, and index; learn to print rows or columns, view data types, missing values, and memory usage.

Learn to index specific rows in pandas dataframes using basic indexing, the lookup function, head and tail, iloc and loc, with start-stop slicing, stepping, and range examples.

Load data from a Postgres server into Python by installing the database connector library, establishing a connection, executing SQL, and creating a pandas data frame.

Index single or multiple columns in a data frame by name, using a list of column names to create a new frame with year, make, and selling price.

Master how to detect and handle duplicated rows and columns in a data frame, using pandas methods like duplicated and drop_duplicates, with inplace options for clean, sanitized data.

Learn to filter data by column names using the data frame filter method, selecting columns like make, model, and transmission, and using like and starting with conditions.

Filter a dataframe using multiple conditions in pandas with and, or, and not operators, applying criteria like Kia, year below 2015, and price below 50,000, using the query method.

Master pandas data type conversion with as_type to turn floats into integers, handle missing values by imputing with the mean, and verify data types in your dataframe.

Convert object data to datetime using pandas to_datetime, extract year, month, day, and time, and handle time zones with UTC to enable date range analysis and filtering.

Explore categorical data in dataframes by computing unique values and value counts, summarizing across all categorical columns with a for loop and ascending/descending options.

Learn to drop single or multiple columns in pandas using drop method, axis and inplace to update the frame or create a new one, including drop any for missing values.

Encode categorical data using a mapping function, then apply label and one-hot encoding while loading marketing data with pandas, identify categorical columns, and map categories to numeric values for ML.

Convert numerical data into categorical bins with pandas cut and labels, then visualize distributions with seaborn to identify central tendency and outliers for informed binning.

Master pandas operations with sorting, single and multi-level, in ascending or descending order. Then learn group by, aggregation, pivoting, and various joins, including concatenations and mergers.

Learn single level sorting in pandas by loading the customer data, inspecting columns, and sorting by country or unit price with optional ascending or descending order.

Utilize the group by function to create single level groupings by country, then aggregate with max, min, sum, mean, median, variance, and standard deviation to reveal country and customer insights.

Explore multi-level grouping by country and customer id, then compute visits, orders, revenue, and aggregated metrics like mean and total per group.

Pivot dataframes in pandas to summarize data by columns, apply aggregation functions, and create multi-level summaries; transpose results and export pivot tables to Excel.

Concatenate two or three pandas data frames using pd.concat to create a final data frame, exploring vertical (axis 0) and horizontal (axis 1) merging with df1, df2, and df3.

Identify and quantify missing values with isnull, manipulate using mean or median, and apply handling techniques to preserve data analysis quality and machine learning algorithms.

Apply context-aware imputation by selecting mean, median, or mode based on data distribution and type. Consider skewness, outliers, and time series patterns, using backward or forward filling when appropriate.

Identify and quantify outliers and anomalies in data to safeguard predictive analytics and forecasting, imputation, and cleaning using visualization, box plots, z-scores, skewness, kurtosis, and the iqr method.

Explore the IQR method to quantify outliers by computing Q1 and Q3 and setting bounds at Q1 minus 1.5 IQR and Q3 plus 1.5 IQR, with deletion, imputation, or capping.

Identify and replace data outliers using the mask function, compare mean and median, and apply median-based imputation to a price column with upper and lower whiskers.