resolve_filter
resolve_filter(
df: pd.DataFrame,
f: list,
rename_map: dict = None
) -> pd.Series
Input DataFrame to filter.
Filter list with operator and operands. Format: [operator, field, value] for basic operations, or [boolean_op, filter1, filter2, ...] for compound filters.
Optional mapping of original to renamed columns.
Boolean Series corresponding to the filter, with True for rows that match the filter criteria.
Supported Operators
Basic Comparison Operators
>, <, >=, <=: Numeric comparisons==, !=: Equality checksisin, notin: Membership tests (value must be a list)contains, contains_case_insensitive: String contains operationsisempty, iszero, iszeroempty: Null/empty checks
Boolean Operators
and, or: Logical AND/ORnand, nor: Negated AND/ORxor, xnor: Exclusive OR and its negationnot: Logical NOT
Examples
# Simple comparison
filter_expensive = resolve_filter(df, [">", "sale_price", 500000])
# Multiple conditions with AND
filter_recent_sales = resolve_filter(df, [
"and",
[">", "sale_year", 2020],
["==", "property_type", "str:SINGLE_FAMILY"]
])
# Using isin for multiple values
filter_neighborhoods = resolve_filter(df, [
"isin", "neighborhood", ["Downtown", "Midtown", "Uptown"]
])
# Complex nested logic
filter_complex = resolve_filter(df, [
"and",
["or",
[">", "bldg_sqft", 2000],
[">", "land_sqft", 10000]
],
["!=", "zoning", "str:INDUSTRIAL"]
])
select_filter
select_filter(df: pd.DataFrame, f: list) -> pd.DataFrame
resolve_filter that returns the filtered DataFrame directly.
Filter expressed as a list. Same format as resolve_filter.
Filtered DataFrame containing only rows that match the filter criteria.
Example
# Select only recent high-value sales
df_filtered = select_filter(df, [
"and",
[">", "sale_year", 2020],
[">", "sale_price", 300000]
])
resolve_bool_filter
resolve_bool_filter(df: pd.DataFrame, f: list) -> pd.Series
List where the first element is the boolean operator and the remaining elements are filter objects.
Boolean Series resulting from applying the boolean operator.
Example
# Combine multiple filters with OR
result = resolve_bool_filter(df, [
"or",
["==", "property_type", "str:CONDO"],
["==", "property_type", "str:TOWNHOUSE"]
])
resolve_not_filter
resolve_not_filter(df: pd.DataFrame, f: list) -> pd.Series
Filter list starting with “not”.
Boolean Series resulting from applying the NOT operator.
Example
# Select everything except vacant properties
result = resolve_not_filter(df, [
"not",
["==", "is_vacant", True]
])
validate_filter
validate_filter(f: list) -> bool
Filter expressed as a list.
True if the filter is valid.
If the value type does not match the operator requirements.
Example
# Valid filter
validate_filter([">", "sale_price", 100000]) # Returns True
# Invalid filter (string value for numeric operator)
validate_filter([">", "sale_price", "expensive"]) # Raises ValueError