Data Analysis Fundamentals final test: 25 free questions

Q1 / 25

How does a DataFrame constructor combine a dictionary of Series with different indexes?

1. AIt combines values strictly by their current positions
2. BIt discards every label absent from the first Series
3. CIt aligns by labels and uses the union of indexes
4. DIt requires all Series labels to be identical first

Q2 / 25

Which window family computes over a moving fixed-count or time-offset partition?

1. AAn exponentially weighted window
2. BAn expanding window
3. CA categorical grouping window
4. DA fixed-count or time-offset rolling window

Q3 / 25

What does the `keys` argument add to a concatenation result?

1. AA left-style restriction to the first index
2. BAn outer MultiIndex level tied to input pieces
3. CA sequential index that erases input labels
4. DAn intersection filter for input columns

Q4 / 25

An analyst must compare two regional measures whose row order differs and must retain regions present in just one input. Which model best meets both needs?

1. AScalar values repeated over the left input positions
2. BUnlabeled arrays paired by their current row positions
3. CSeries converted to arrays before their labels are compared
4. DLabeled Series aligned over the union of region labels

Q5 / 25

Identifiers look like `region-team-member`, but the region and team may themselves contain hyphens. The analyst must isolate only the final member token and retain the unsplit prefix as one field. Which operation meets both constraints?

1. A`s.str.rsplit("-", n=1, expand=True)`
2. B`s.str.split("-", n=1, expand=True)`
3. C`s.str.replace("-", "", regex=False)`
4. D`s.str.split("-", expand=True)`

Q6 / 25

A table has columns `store`, `day`, and `sales`. Each observation is identified by the pair `(store, day)`, and both identifiers must remain available as data columns. Which setup fits?

1. A`df.set_index(['store', 'day'], drop=False)`
2. B`df.drop_duplicates(['store', 'day'])`
3. C`df.reset_index(['store', 'day'], drop=False)`
4. D`df.set_index('store', drop=True)`

Q7 / 25

How can a Parquet reader avoid materializing unneeded columns?

1. APass the required names through `columns`
2. BPass the required names through `partition_cols`
3. CPass the required names through `compression`
4. DPass the required names through `index`

Q8 / 25

What index condition applies when a rolling window is defined by a time offset?

1. AThe corresponding time-based index must be monotonic
2. BThe corresponding time-based index must contain unique days
3. CThe time-based index must have equal intervals
4. DThe corresponding time-based index must be categorical

Q9 / 25

What does `validate="one_to_one"` check before a merge?

1. AThat input indexes become result columns
2. BThat merge keys are unique on both sides
3. CThat overlapping value columns have suffixes
4. DThat unmatched keys exist on both sides

Q10 / 25

Monthly `PeriodIndex` data must become daily-resolution timestamps placed at each span's last representable point. Which conversion expresses both choices?

1. AUse `to_timestamp('D', how='end')`
2. BUse `asfreq('D', how='start')` and keep periods
3. CUse `to_timestamp('D', how='start')`
4. DUse `to_period('D', how='end')`

Q11 / 25

Which representation is a documented candidate for low-cardinality text currently stored as `object`?

1. AUse `timedelta64[ns]` values
2. BUse `complex128` numbers
3. CUse `datetime64[ns]` values
4. DConvert the column to `category`

Q12 / 25

Which selection approaches appear in the beginner pandas workflow?

1. ASelection by join engine, file writer, and group output
2. BSelection by plot color, export format, and workbook name
3. CSelection by package alias, scalar dtype, and chart legend
4. DSelection by label, by position, and by Boolean condition

Q13 / 25

A financial table must highlight negative cells conditionally and retain that cell styling in an Excel file. The design currently uses external classes attached with `.set_td_classes()`. What change satisfies both requirements?

1. AMove the negative-value rule into a table-wide hover pseudo-selector
2. BFormat the values as strings with `.format()` and export the same CSS classes
3. CKeep `.set_td_classes()` and add the classes through `.set_table_styles()`
4. DMap the conditional cell properties with `.map()` before Excel export

Q14 / 25

What happens to an input value that is absent from an explicitly declared category set?

1. AIt is appended to the category set automatically
2. BIt is preserved as an unordered object value in the same column
3. CIt replaces first declared category
4. DIt becomes a missing categorical value

Q15 / 25

Inside `DataFrame.query()`, an index and a column are both named `a`. What does the expression `a > 2` use?

1. AThe index named `a`
2. BThe column named `a`
3. CBoth objects combined positionally
4. DNeither object because the name is rejected

Q16 / 25

A pipe-delimited file has no header, `NA?` is its sole missing marker, and `account_id` must remain text. Only three named columns are needed. Which ingestion design satisfies all constraints?

1. ASet `sep="|"`, explicit `names`, `header=None`, `usecols`, text dtype for `account_id`, `na_values=["NA?"]`, and `keep_default_na=False`
2. BUse `read_json(orient="records")` and treat each pipe-delimited line as a JSON object
3. CSet `header=0`, keep every column, and rely on the default missing-marker set
4. DUse default `read_csv()` inference, then rename columns and convert all values to floats

Q17 / 25

What happens if a string passed to `groupby()` matches both a column name and an index-level name?

1. Apandas raises a `ValueError` because the key is ambiguous
2. Bpandas groups by the column because columns take precedence
3. Cpandas combines both matches into a two-key grouping
4. Dpandas groups by the index because levels take precedence

Q18 / 25

A Series represents one observation and its index matches the target DataFrame columns. How should it be appended as a row?

1. AConvert it to a DataFrame and concatenate without transposing
2. BConvert it to a DataFrame, transpose it, then concatenate
3. CDiscard its index and use the values as provenance keys
4. DConcatenate it directly with `axis=1` and preserve labels

Q19 / 25

How should a datetime-like Series named `events` expose its hour values without scalar iteration?

1. AUse `events.hour.dt`
2. BUse `events.dt.hour`
3. CUse `events.astype('object').hour`
4. DUse `events.index.hour`

Q20 / 25

Which pandas operation is associated with the documented thread-safety concern?

1. AComparing one local Series with a scalar
2. BCopying DataFrames shared among threads
3. CChecking one local DataFrame with `empty`
4. DReducing one local Boolean Series with `any()`

Q21 / 25

Which pairing correctly matches a workflow need to a pandas capability?

1. AChange table layout with read_csv(); change time frequency with concat()
2. BChange table layout with plot(); change time frequency with Index
3. CChange table layout with isna(); change time frequency with columns
4. DChange table layout with reshaping; change time frequency with resampling

Q22 / 25

Which pandas plotting arguments map two named DataFrame columns to the horizontal and vertical axes?

1. APass the column names through `layout` and `subplots`
2. BPass columns through `xerr` and `yerr`
3. CPass the column names through `style` and `label`
4. DPass the column names through `x` and `y`

Q23 / 25

A transformation is already expressible as an array operation. Which implementation best matches the documented decision rule?

1. AUse `select_dtypes()` as the computation instead of applying an operation
2. BUse the direct vectorized expression because it clearly expresses the operation
3. CConvert labels to rows and invoke `reindex()` for each value
4. DWrap every scalar in `DataFrame.map()` before trying the array operation

Q24 / 25

Why is `df['foo'].replace(1, 5, inplace=True)` unsafe under Copy-on-Write?

1. AIt invokes an inplace method on an intermediate selected column
2. B`replace()` can operate only on whole DataFrames
3. CThe column selector automatically removes `foo` from `df`
4. DCopy-on-Write forbids replacing numeric values

Q25 / 25

What is the difference between `engine="pyarrow"` and `dtype_backend="pyarrow"` on a supported reader?

1. AThe engine chooses parsing; the dtype backend chooses Arrow-backed result types
2. BThe engine writes Parquet; the dtype backend writes JSON records
3. CThe engine selects SQL queries; the dtype backend selects database tables
4. DThe engine chooses returned row and column labels; the dtype backend chooses which source rows are read