Examius
🇮🇳 Data Science & Machine Learning · flashcards
Data Science & Machine Learning Exploratory Data Analysis and Visualization Flashcards
50 question-and-answer cards covering Exploratory Data Analysis and Visualization as it is examined in Data Science & Machine Learning. 24 of them are printed below, taken from across the deck — no signup, no paywall on the preview.
24 sample cards from the Exploratory Data Analysis and Visualization deck
Sampled from the end of the deck, so these are different cards from the ones shown on the syllabus page.
What does each part of a box plot represent?
The box spans Q1 to Q3 (the IQR), the line inside is the median (Q2), the whiskers extend to the most extreme points within 1.5×IQR, and points beyond the whiskers are plotted as outliers.
What is a histogram bin, and why does bin width matter?
A bin is an interval into which data values are grouped; the bar height is the count in that interval. Bin width controls granularity: too wide hides structure, too narrow adds noise; choice strongly affects the histogram's appearance.
What is a kernel density estimate (KDE) plot?
A smoothed, continuous estimate of a variable's probability density function, produced by summing kernels placed at each data point — a smooth alternative to a histogram.
What is a violin plot and what does it add over a box plot?
A violin plot combines a box plot with a mirrored KDE, showing not just the quartiles and median but also the full shape (density) of the distribution, including multimodality.
In Matplotlib, how do you add axis labels, a title, and a legend?
Use ax.set_xlabel(), ax.set_ylabel(), ax.set_title(), and ax.legend() (the legend uses the 'label' argument given to each plotted series).
What is a colormap (cmap) in plotting?
A mapping from data values to colors used to encode a numeric dimension. Types include sequential (ordered data), diverging (data with a meaningful midpoint), and qualitative (categorical data).
How can you change the size of a Matplotlib figure?
Pass figsize=(width, height) in inches to plt.figure() or plt.subplots(), e.g., plt.subplots(figsize=(10, 6)).
What is Seaborn and how does it relate to Matplotlib?
Seaborn is a high-level statistical data visualization library built on top of Matplotlib. It provides attractive default styles and concise functions for complex statistical plots, integrating closely with pandas DataFrames.
What is the difference between Seaborn's figure-level and axes-level functions?
Axes-level functions (e.g., sns.scatterplot, sns.histplot) draw onto a single Matplotlib Axes; figure-level functions (e.g., sns.relplot, sns.displot, sns.catplot) manage their own Figure and can create faceted grids of multiple subplots.
Which Seaborn function creates distribution plots, and what kinds can it show?
sns.displot (figure-level) and sns.histplot/sns.kdeplot/sns.ecdfplot (axes-level) create distribution plots — histograms, KDE curves, and empirical cumulative distribution functions (ECDFs).
What does an ECDF (empirical cumulative distribution function) plot show?
It plots, for each value, the proportion of data points less than or equal to that value — a monotonically increasing step curve from 0 to 1 that shows percentiles without binning.
Which Seaborn function is used for relationship plots between two numerical variables?
sns.relplot (figure-level) with kind='scatter' or kind='line', or the axes-level sns.scatterplot and sns.lineplot.
How do hue, size, and style parameters enrich a Seaborn relationship plot?
They add extra dimensions via semantic mapping: hue maps a variable to color, size maps a variable to marker/line size, and style maps a variable to marker shape or line dash — enabling multivariate visualization in one plot.
What is a regression plot (sns.regplot / lmplot) used for?
It draws a scatter plot of two numerical variables and overlays a fitted regression line with a confidence interval, to visualize a linear (or polynomial) trend between them.
What categorical plot types does Seaborn's catplot support?
strip, swarm (scatter-type); box, violin, boxen (distribution-type); and bar, count, point (estimate-type) plots — selected via the kind parameter.
What is the difference between a strip plot and a swarm plot?
Both show individual points across categories; a strip plot may overlap points (often with jitter), while a swarm plot adjusts point positions so none overlap, giving a clearer view of density at small data sizes.
What does a Seaborn count plot show?
It shows the number of observations in each category (a bar chart of frequencies), equivalent to a histogram for a single categorical variable.
What does the bar in a Seaborn barplot represent by default, and what do the error bars show?
By default the bar height is the mean of the numerical variable for each category, and the error bars show a 95% confidence interval of that estimate (via bootstrapping).
What is a heatmap and a common EDA use for it?
A heatmap encodes matrix values as colors. A common EDA use is visualizing a correlation matrix to quickly see which variable pairs are strongly positively or negatively correlated.
What is Plotly and what key feature distinguishes it from Matplotlib/Seaborn?
Plotly is a graphing library that produces interactive, web-based visualizations supporting hover tooltips, zooming, panning, and toggling — unlike the static images typically produced by Matplotlib and Seaborn.
What is Plotly Express?
plotly.express (imported as px) is the high-level, concise interface to Plotly that creates entire interactive figures with a single function call (e.g., px.scatter, px.line, px.bar) directly from a DataFrame.
What are the principles of effective insight communication in data visualization?
Choose the right chart for the message, reduce clutter (high data-ink ratio), use clear titles and labels, highlight the key takeaway, use color purposefully, and tailor the visualization to the audience.
What is a dashboard in the context of data reporting?
A single interface that consolidates and displays multiple related visualizations and key metrics (KPIs), often interactive, to monitor and communicate the state of a system or business at a glance.
Name common Python tools for building interactive data dashboards.
Dash (by Plotly), Streamlit, Panel, and Voila — used to turn Python analyses and Plotly/Matplotlib charts into shareable interactive web dashboards.
What this deck covers
The Exploratory Data Analysis and Visualization deck follows the Data Science & Machine Learning Exploratory Data Analysis and Visualization syllabus — 4 chapters and 12 topics — so questions land on material that is genuinely examinable rather than trivia around it. That works out to roughly 12.5 cards per chapter.
Answers are written to be recallable, not just readable — averaging about 183 characters, which is long enough to carry the reasoning and short enough to say out loud.
A deck like this earns its keep on the second and third pass. Read the syllabus first so you know the shape of the subject, then use the cards to find the specific facts that have not stuck.
Exploratory Data Analysis and Visualization flashcards FAQ
How many Exploratory Data Analysis and Visualization flashcards are in this Data Science & Machine Learning deck?
50 cards. This page previews 24 of them, sampled evenly across the deck so you can judge the difficulty before installing anything.
Are these Data Science & Machine Learning flashcards free?
Yes. The preview here is free to read with no signup, and the full 50-card deck is free inside the Examius app.
What do the Exploratory Data Analysis and Visualization cards cover?
They follow the Data Science & Machine Learning Exploratory Data Analysis and Visualization syllabus — 4 chapters and 12 topics — so the questions track what is actually examinable.
How should I use these flashcards?
Read the syllabus first so you know the shape of the subject, then drill the deck. Examius schedules each card with spaced repetition, so cards you keep missing come back sooner and ones you know drift further apart.