Svep för att visa menyn

tracemalloc and memory_profiler show how much memory is used. objgraph shows what is in memory – how many instances of each type exist, what references what, and where reference cycles are hiding. It is the go-to tool for diagnosing memory leaks.

Installation

pip install objgraph

Counting Objects by Type

objgraph.show_most_common_types() lists the types with the most live instances:


              123456789101112
            
import objgraph

# Creating some objects and checking what's in memory
class Invoice:
    def __init__(self, invoice_id, amount):
        self.invoice_id = invoice_id
        self.amount = amount

invoices = [Invoice(invoice_id, invoice_id * 250.0) for invoice_id in range(1000)]
cached_results = {"q1": list(range(500)), "q2": list(range(500))}

objgraph.show_most_common_types(limit=10)

This gives an instant overview of what is consuming memory.

Detecting Growth

objgraph.show_growth() compares the current object counts to the previous call and shows only what grew:


              1234567891011121314
            
import objgraph

class ReportSection:
    def __init__(self, section_id):
        self.section_id = section_id

objgraph.show_growth()  # Baseline snapshot

# Simulating a function that leaks objects
leaked_sections = []
for section_id in range(500):
    leaked_sections.append(ReportSection(section_id))

objgraph.show_growth()  # Shows what grew since the baseline

If show_growth() keeps showing the same type growing on every call, that type is leaking.

Finding Reference Cycles

objgraph.find_backref_chain() traces the reference path from a specific object back to a root – showing exactly what is keeping it alive:


              12345678910111213141516171819202122
            
import objgraph
import gc

class Node:
    def __init__(self, label):
        self.label = label
        self.child = None

# Creating a cycle
node_a = Node("parent")
node_b = Node("child")
node_a.child = node_b
node_b.child = node_a  # Cycle

# Finding what keeps node_a alive
chain = objgraph.find_backref_chain(
    node_a,
    objgraph.is_proper_module
)
print(chain)

gc.collect()  # Collecting the cycle after inspection

Tracking a Specific Type

objgraph.by_type() returns all live instances of a given type – useful for checking that a factory or cache is not accumulating objects:


              12345678910111213
            
import objgraph

class DataChunk:
    def __init__(self, chunk_id):
        self.chunk_id = chunk_id

chunks = [DataChunk(chunk_id) for chunk_id in range(5)]

live_chunks = objgraph.by_type("DataChunk")
print(f"Live DataChunk instances: {len(live_chunks)}")

for chunk in live_chunks:
    print(f"  chunk_id={chunk.chunk_id}")

`objgraph` vs `tracemalloc`

Var allt tydligt?

Tack för dina kommentarer!

Avsnitt 3. Kapitel 3

Fråga AI

Fråga vad du vill eller prova någon av de föreslagna frågorna för att starta vårt samtal

Visualizing Object Graphs with objgraph

Installation

pip install objgraph

Counting Objects by Type

objgraph.show_most_common_types() lists the types with the most live instances:


              123456789101112
            
import objgraph

# Creating some objects and checking what's in memory
class Invoice:
    def __init__(self, invoice_id, amount):
        self.invoice_id = invoice_id
        self.amount = amount

invoices = [Invoice(invoice_id, invoice_id * 250.0) for invoice_id in range(1000)]
cached_results = {"q1": list(range(500)), "q2": list(range(500))}

objgraph.show_most_common_types(limit=10)

This gives an instant overview of what is consuming memory.

Detecting Growth

objgraph.show_growth() compares the current object counts to the previous call and shows only what grew:


              1234567891011121314
            
import objgraph

class ReportSection:
    def __init__(self, section_id):
        self.section_id = section_id

objgraph.show_growth()  # Baseline snapshot

# Simulating a function that leaks objects
leaked_sections = []
for section_id in range(500):
    leaked_sections.append(ReportSection(section_id))

objgraph.show_growth()  # Shows what grew since the baseline

If show_growth() keeps showing the same type growing on every call, that type is leaking.

Finding Reference Cycles

objgraph.find_backref_chain() traces the reference path from a specific object back to a root – showing exactly what is keeping it alive:


              12345678910111213141516171819202122
            
import objgraph
import gc

class Node:
    def __init__(self, label):
        self.label = label
        self.child = None

# Creating a cycle
node_a = Node("parent")
node_b = Node("child")
node_a.child = node_b
node_b.child = node_a  # Cycle

# Finding what keeps node_a alive
chain = objgraph.find_backref_chain(
    node_a,
    objgraph.is_proper_module
)
print(chain)

gc.collect()  # Collecting the cycle after inspection

Tracking a Specific Type

objgraph.by_type() returns all live instances of a given type – useful for checking that a factory or cache is not accumulating objects:


              12345678910111213
            
import objgraph

class DataChunk:
    def __init__(self, chunk_id):
        self.chunk_id = chunk_id

chunks = [DataChunk(chunk_id) for chunk_id in range(5)]

live_chunks = objgraph.by_type("DataChunk")
print(f"Live DataChunk instances: {len(live_chunks)}")

for chunk in live_chunks:
    print(f"  chunk_id={chunk.chunk_id}")

`objgraph` vs `tracemalloc`

Var allt tydligt?

Tack för dina kommentarer!

Avsnitt 3. Kapitel 3

Visualizing Object Graphs with objgraph

Installation

Counting Objects by Type

Detecting Growth

Finding Reference Cycles

Tracking a Specific Type

objgraph vs tracemalloc

Visualizing Object Graphs with objgraph

Installation

Counting Objects by Type

Detecting Growth

Finding Reference Cycles

Tracking a Specific Type

objgraph vs tracemalloc

`objgraph` vs `tracemalloc`

`objgraph` vs `tracemalloc`