Python | Pandas TimedeltaIndex.take (): StackOverflow Questions

How can I make a time delay in Python?

I would like to know how to put a time delay in a Python script.

How to delete a file or folder in Python?

How do I delete a file or folder in Python?

Difference between del, remove, and pop on lists

>>> a=[1,2,3]
>>> a.remove(2)
>>> a
[1, 3]
>>> a=[1,2,3]
>>> del a[1]
>>> a
[1, 3]
>>> a= [1,2,3]
>>> a.pop(1)
2
>>> a
[1, 3]
>>> 

Is there any difference between the above three methods to remove an element from a list?

Is there a simple way to delete a list element by value?

I want to remove a value from a list if it exists in the list (which it may not).

a = [1, 2, 3, 4]
b = a.index(6)

del a[b]
print(a)

The above case (in which it does not exist) shows the following error:

Traceback (most recent call last):
  File "D:zjm_codea.py", line 6, in <module>
    b = a.index(6)
ValueError: list.index(x): x not in list

So I have to do this:

a = [1, 2, 3, 4]

try:
    b = a.index(6)
    del a[b]
except:
    pass

print(a)

But is there not a simpler way to do this?

How do I remove/delete a folder that is not empty?

Question by Amara

I am getting an "access is denied" error when I attempt to delete a folder that is not empty. I used the following command in my attempt: os.remove("/folder_name").

What is the most effective way of removing/deleting a folder/directory that is not empty?

Split Strings into words with multiple word boundary delimiters

I think what I want to do is a fairly common task but I"ve found no reference on the web. I have text with punctuation, and I want a list of the words.

"Hey, you - what are you doing here!?"

should be

["hey", "you", "what", "are", "you", "doing", "here"]

But Python"s str.split() only works with one argument, so I have all words with the punctuation after I split with whitespace. Any ideas?

Split string with multiple delimiters in Python

I found some answers online, but I have no experience with regular expressions, which I believe is what is needed here.

I have a string that needs to be split by either a ";" or ", " That is, it has to be either a semicolon or a comma followed by a space. Individual commas without trailing spaces should be left untouched

Example string:

"b-staged divinylsiloxane-bis-benzocyclobutene [124221-30-3], mesitylene [000108-67-8]; polymerized 1,2-dihydro-2,2,4- trimethyl quinoline [026780-96-1]"

should be split into a list containing the following:

("b-staged divinylsiloxane-bis-benzocyclobutene [124221-30-3]" , "mesitylene [000108-67-8]", "polymerized 1,2-dihydro-2,2,4- trimethyl quinoline [026780-96-1]") 

How to save/restore a model after training?

After you train a model in Tensorflow:

1. How do you save the trained model?
2. How do you later restore this saved model?

How to delete the contents of a folder?

Question by Unkwntech

How can I delete the contents of a local folder in Python?

The current project is for Windows, but I would like to see *nix also.

How do I remove/delete a virtualenv?

I created an environment with the following command: virtualenv venv --distribute

I cannot remove it with the following command: rmvirtualenv venv - This is part of virtualenvwrapper as mentioned in answer below for virtualenvwrapper

I do an lson my current directory and I still see venv

The only way I can remove it seems to be: sudo rm -rf venv

Note that the environment is not active. I"m running Ubuntu 11.10. Any ideas? I"ve tried rebooting my system to no avail.

Answer #1

The Python 3 range() object doesn"t produce numbers immediately; it is a smart sequence object that produces numbers on demand. All it contains is your start, stop and step values, then as you iterate over the object the next integer is calculated each iteration.

The object also implements the object.__contains__ hook, and calculates if your number is part of its range. Calculating is a (near) constant time operation ^*. There is never a need to scan through all possible integers in the range.

From the range() object documentation:

The advantage of the range type over a regular list or tuple is that a range object will always take the same (small) amount of memory, no matter the size of the range it represents (as it only stores the start, stop and step values, calculating individual items and subranges as needed).

So at a minimum, your range() object would do:

class my_range:
    def __init__(self, start, stop=None, step=1, /):
        if stop is None:
            start, stop = 0, start
        self.start, self.stop, self.step = start, stop, step
        if step < 0:
            lo, hi, step = stop, start, -step
        else:
            lo, hi = start, stop
        self.length = 0 if lo > hi else ((hi - lo - 1) // step) + 1

    def __iter__(self):
        current = self.start
        if self.step < 0:
            while current > self.stop:
                yield current
                current += self.step
        else:
            while current < self.stop:
                yield current
                current += self.step

    def __len__(self):
        return self.length

    def __getitem__(self, i):
        if i < 0:
            i += self.length
        if 0 <= i < self.length:
            return self.start + i * self.step
        raise IndexError("my_range object index out of range")

    def __contains__(self, num):
        if self.step < 0:
            if not (self.stop < num <= self.start):
                return False
        else:
            if not (self.start <= num < self.stop):
                return False
        return (num - self.start) % self.step == 0

This is still missing several things that a real range() supports (such as the .index() or .count() methods, hashing, equality testing, or slicing), but should give you an idea.

I also simplified the __contains__ implementation to only focus on integer tests; if you give a real range() object a non-integer value (including subclasses of int), a slow scan is initiated to see if there is a match, just as if you use a containment test against a list of all the contained values. This was done to continue to support other numeric types that just happen to support equality testing with integers but are not expected to support integer arithmetic as well. See the original Python issue that implemented the containment test.

---

* Near constant time because Python integers are unbounded and so math operations also grow in time as N grows, making this a O(log N) operation. Since it’s all executed in optimised C code and Python stores integer values in 30-bit chunks, you’d run out of memory before you saw any performance impact due to the size of the integers involved here.

Answer #2

You have four main options for converting types in pandas:

1. to_numeric() - provides functionality to safely convert non-numeric types (e.g. strings) to a suitable numeric type. (See also to_datetime() and to_timedelta().)

2. astype() - convert (almost) any type to (almost) any other type (even if it"s not necessarily sensible to do so). Also allows you to convert to categorial types (very useful).

3. infer_objects() - a utility method to convert object columns holding Python objects to a pandas type if possible.

4. convert_dtypes() - convert DataFrame columns to the "best possible" dtype that supports pd.NA (pandas" object to indicate a missing value).

Read on for more detailed explanations and usage of each of these methods.

---

1. to_numeric()

The best way to convert one or more columns of a DataFrame to numeric values is to use pandas.to_numeric().

This function will try to change non-numeric objects (such as strings) into integers or floating point numbers as appropriate.

Basic usage

The input to to_numeric() is a Series or a single column of a DataFrame.

>>> s = pd.Series(["8", 6, "7.5", 3, "0.9"]) # mixed string and numeric values
>>> s
0      8
1      6
2    7.5
3      3
4    0.9
dtype: object

>>> pd.to_numeric(s) # convert everything to float values
0    8.0
1    6.0
2    7.5
3    3.0
4    0.9
dtype: float64

As you can see, a new Series is returned. Remember to assign this output to a variable or column name to continue using it:

# convert Series
my_series = pd.to_numeric(my_series)

# convert column "a" of a DataFrame
df["a"] = pd.to_numeric(df["a"])

You can also use it to convert multiple columns of a DataFrame via the apply() method:

# convert all columns of DataFrame
df = df.apply(pd.to_numeric) # convert all columns of DataFrame

# convert just columns "a" and "b"
df[["a", "b"]] = df[["a", "b"]].apply(pd.to_numeric)

As long as your values can all be converted, that"s probably all you need.

Error handling

But what if some values can"t be converted to a numeric type?

to_numeric() also takes an errors keyword argument that allows you to force non-numeric values to be NaN, or simply ignore columns containing these values.

Here"s an example using a Series of strings s which has the object dtype:

>>> s = pd.Series(["1", "2", "4.7", "pandas", "10"])
>>> s
0         1
1         2
2       4.7
3    pandas
4        10
dtype: object

The default behaviour is to raise if it can"t convert a value. In this case, it can"t cope with the string "pandas":

>>> pd.to_numeric(s) # or pd.to_numeric(s, errors="raise")
ValueError: Unable to parse string

Rather than fail, we might want "pandas" to be considered a missing/bad numeric value. We can coerce invalid values to NaN as follows using the errors keyword argument:

>>> pd.to_numeric(s, errors="coerce")
0     1.0
1     2.0
2     4.7
3     NaN
4    10.0
dtype: float64

The third option for errors is just to ignore the operation if an invalid value is encountered:

>>> pd.to_numeric(s, errors="ignore")
# the original Series is returned untouched

This last option is particularly useful when you want to convert your entire DataFrame, but don"t not know which of our columns can be converted reliably to a numeric type. In that case just write:

df.apply(pd.to_numeric, errors="ignore")

The function will be applied to each column of the DataFrame. Columns that can be converted to a numeric type will be converted, while columns that cannot (e.g. they contain non-digit strings or dates) will be left alone.

Downcasting

By default, conversion with to_numeric() will give you either a int64 or float64 dtype (or whatever integer width is native to your platform).

That"s usually what you want, but what if you wanted to save some memory and use a more compact dtype, like float32, or int8?

to_numeric() gives you the option to downcast to either "integer", "signed", "unsigned", "float". Here"s an example for a simple series s of integer type:

>>> s = pd.Series([1, 2, -7])
>>> s
0    1
1    2
2   -7
dtype: int64

Downcasting to "integer" uses the smallest possible integer that can hold the values:

>>> pd.to_numeric(s, downcast="integer")
0    1
1    2
2   -7
dtype: int8

Downcasting to "float" similarly picks a smaller than normal floating type:

>>> pd.to_numeric(s, downcast="float")
0    1.0
1    2.0
2   -7.0
dtype: float32

---

2. astype()

The astype() method enables you to be explicit about the dtype you want your DataFrame or Series to have. It"s very versatile in that you can try and go from one type to the any other.

Basic usage

Just pick a type: you can use a NumPy dtype (e.g. np.int16), some Python types (e.g. bool), or pandas-specific types (like the categorical dtype).

Call the method on the object you want to convert and astype() will try and convert it for you:

# convert all DataFrame columns to the int64 dtype
df = df.astype(int)

# convert column "a" to int64 dtype and "b" to complex type
df = df.astype({"a": int, "b": complex})

# convert Series to float16 type
s = s.astype(np.float16)

# convert Series to Python strings
s = s.astype(str)

# convert Series to categorical type - see docs for more details
s = s.astype("category")

Notice I said "try" - if astype() does not know how to convert a value in the Series or DataFrame, it will raise an error. For example if you have a NaN or inf value you"ll get an error trying to convert it to an integer.

As of pandas 0.20.0, this error can be suppressed by passing errors="ignore". Your original object will be return untouched.

Be careful

astype() is powerful, but it will sometimes convert values "incorrectly". For example:

>>> s = pd.Series([1, 2, -7])
>>> s
0    1
1    2
2   -7
dtype: int64

These are small integers, so how about converting to an unsigned 8-bit type to save memory?

>>> s.astype(np.uint8)
0      1
1      2
2    249
dtype: uint8

The conversion worked, but the -7 was wrapped round to become 249 (i.e. 2⁸ - 7)!

Trying to downcast using pd.to_numeric(s, downcast="unsigned") instead could help prevent this error.

---

3. infer_objects()

Version 0.21.0 of pandas introduced the method infer_objects() for converting columns of a DataFrame that have an object datatype to a more specific type (soft conversions).

For example, here"s a DataFrame with two columns of object type. One holds actual integers and the other holds strings representing integers:

>>> df = pd.DataFrame({"a": [7, 1, 5], "b": ["3","2","1"]}, dtype="object")
>>> df.dtypes
a    object
b    object
dtype: object

Using infer_objects(), you can change the type of column "a" to int64:

>>> df = df.infer_objects()
>>> df.dtypes
a     int64
b    object
dtype: object

Column "b" has been left alone since its values were strings, not integers. If you wanted to try and force the conversion of both columns to an integer type, you could use df.astype(int) instead.

---

4. convert_dtypes()

Version 1.0 and above includes a method convert_dtypes() to convert Series and DataFrame columns to the best possible dtype that supports the pd.NA missing value.

Here "best possible" means the type most suited to hold the values. For example, this a pandas integer type if all of the values are integers (or missing values): an object column of Python integer objects is converted to Int64, a column of NumPy int32 values will become the pandas dtype Int32.

With our object DataFrame df, we get the following result:

>>> df.convert_dtypes().dtypes                                             
a     Int64
b    string
dtype: object

Since column "a" held integer values, it was converted to the Int64 type (which is capable of holding missing values, unlike int64).

Column "b" contained string objects, so was changed to pandas" string dtype.

By default, this method will infer the type from object values in each column. We can change this by passing infer_objects=False:

>>> df.convert_dtypes(infer_objects=False).dtypes                          
a    object
b    string
dtype: object

Now column "a" remained an object column: pandas knows it can be described as an "integer" column (internally it ran infer_dtype) but didn"t infer exactly what dtype of integer it should have so did not convert it. Column "b" was again converted to "string" dtype as it was recognised as holding "string" values.

Answer #3

Since this question was asked in 2010, there has been real simplification in how to do simple multithreading with Python with map and pool.

The code below comes from an article/blog post that you should definitely check out (no affiliation) - Parallelism in one line: A Better Model for Day to Day Threading Tasks. I"ll summarize below - it ends up being just a few lines of code:

from multiprocessing.dummy import Pool as ThreadPool
pool = ThreadPool(4)
results = pool.map(my_function, my_array)

Which is the multithreaded version of:

results = []
for item in my_array:
    results.append(my_function(item))

---

Description

Map is a cool little function, and the key to easily injecting parallelism into your Python code. For those unfamiliar, map is something lifted from functional languages like Lisp. It is a function which maps another function over a sequence.

Map handles the iteration over the sequence for us, applies the function, and stores all of the results in a handy list at the end.

---

Implementation

Parallel versions of the map function are provided by two libraries:multiprocessing, and also its little known, but equally fantastic step child:multiprocessing.dummy.

multiprocessing.dummy is exactly the same as multiprocessing module, but uses threads instead (an important distinction - use multiple processes for CPU-intensive tasks; threads for (and during) I/O):

multiprocessing.dummy replicates the API of multiprocessing, but is no more than a wrapper around the threading module.

import urllib2
from multiprocessing.dummy import Pool as ThreadPool

urls = [
  "http://www.python.org",
  "http://www.python.org/about/",
  "http://www.onlamp.com/pub/a/python/2003/04/17/metaclasses.html",
  "http://www.python.org/doc/",
  "http://www.python.org/download/",
  "http://www.python.org/getit/",
  "http://www.python.org/community/",
  "https://wiki.python.org/moin/",
]

# Make the Pool of workers
pool = ThreadPool(4)

# Open the URLs in their own threads
# and return the results
results = pool.map(urllib2.urlopen, urls)

# Close the pool and wait for the work to finish
pool.close()
pool.join()

And the timing results:

Single thread:   14.4 seconds
       4 Pool:   3.1 seconds
       8 Pool:   1.4 seconds
      13 Pool:   1.3 seconds

---

Passing multiple arguments (works like this only in Python 3.3 and later):

To pass multiple arrays:

results = pool.starmap(function, zip(list_a, list_b))

Or to pass a constant and an array:

results = pool.starmap(function, zip(itertools.repeat(constant), list_a))

If you are using an earlier version of Python, you can pass multiple arguments via this workaround).

(Thanks to user136036 for the helpful comment.)

Answer #4

In Python, what is the purpose of __slots__ and what are the cases one should avoid this?

TLDR:

The special attribute __slots__ allows you to explicitly state which instance attributes you expect your object instances to have, with the expected results:

1. faster attribute access.
2. space savings in memory.

The space savings is from

1. Storing value references in slots instead of __dict__.
2. Denying __dict__ and __weakref__ creation if parent classes deny them and you declare __slots__.

Quick Caveats

Small caveat, you should only declare a particular slot one time in an inheritance tree. For example:

class Base:
    __slots__ = "foo", "bar"

class Right(Base):
    __slots__ = "baz", 

class Wrong(Base):
    __slots__ = "foo", "bar", "baz"        # redundant foo and bar

Python doesn"t object when you get this wrong (it probably should), problems might not otherwise manifest, but your objects will take up more space than they otherwise should. Python 3.8:

>>> from sys import getsizeof
>>> getsizeof(Right()), getsizeof(Wrong())
(56, 72)

This is because the Base"s slot descriptor has a slot separate from the Wrong"s. This shouldn"t usually come up, but it could:

>>> w = Wrong()
>>> w.foo = "foo"
>>> Base.foo.__get__(w)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: foo
>>> Wrong.foo.__get__(w)
"foo"

The biggest caveat is for multiple inheritance - multiple "parent classes with nonempty slots" cannot be combined.

To accommodate this restriction, follow best practices: Factor out all but one or all parents" abstraction which their concrete class respectively and your new concrete class collectively will inherit from - giving the abstraction(s) empty slots (just like abstract base classes in the standard library).

See section on multiple inheritance below for an example.

Requirements:

• To have attributes named in __slots__ to actually be stored in slots instead of a __dict__, a class must inherit from object (automatic in Python 3, but must be explicit in Python 2).

• To prevent the creation of a __dict__, you must inherit from object and all classes in the inheritance must declare __slots__ and none of them can have a "__dict__" entry.

There are a lot of details if you wish to keep reading.

Why use __slots__: Faster attribute access.

The creator of Python, Guido van Rossum, states that he actually created __slots__ for faster attribute access.

It is trivial to demonstrate measurably significant faster access:

import timeit

class Foo(object): __slots__ = "foo",

class Bar(object): pass

slotted = Foo()
not_slotted = Bar()

def get_set_delete_fn(obj):
    def get_set_delete():
        obj.foo = "foo"
        obj.foo
        del obj.foo
    return get_set_delete

and

>>> min(timeit.repeat(get_set_delete_fn(slotted)))
0.2846834529991611
>>> min(timeit.repeat(get_set_delete_fn(not_slotted)))
0.3664822799983085

The slotted access is almost 30% faster in Python 3.5 on Ubuntu.

>>> 0.3664822799983085 / 0.2846834529991611
1.2873325658284342

In Python 2 on Windows I have measured it about 15% faster.

Why use __slots__: Memory Savings

Another purpose of __slots__ is to reduce the space in memory that each object instance takes up.

My own contribution to the documentation clearly states the reasons behind this:

The space saved over using __dict__ can be significant.

SQLAlchemy attributes a lot of memory savings to __slots__.

To verify this, using the Anaconda distribution of Python 2.7 on Ubuntu Linux, with guppy.hpy (aka heapy) and sys.getsizeof, the size of a class instance without __slots__ declared, and nothing else, is 64 bytes. That does not include the __dict__. Thank you Python for lazy evaluation again, the __dict__ is apparently not called into existence until it is referenced, but classes without data are usually useless. When called into existence, the __dict__ attribute is a minimum of 280 bytes additionally.

In contrast, a class instance with __slots__ declared to be () (no data) is only 16 bytes, and 56 total bytes with one item in slots, 64 with two.

For 64 bit Python, I illustrate the memory consumption in bytes in Python 2.7 and 3.6, for __slots__ and __dict__ (no slots defined) for each point where the dict grows in 3.6 (except for 0, 1, and 2 attributes):

       Python 2.7             Python 3.6
attrs  __slots__  __dict__*   __slots__  __dict__* | *(no slots defined)
none   16         56 + 272†   16         56 + 112† | †if __dict__ referenced
one    48         56 + 272    48         56 + 112
two    56         56 + 272    56         56 + 112
six    88         56 + 1040   88         56 + 152
11     128        56 + 1040   128        56 + 240
22     216        56 + 3344   216        56 + 408     
43     384        56 + 3344   384        56 + 752

So, in spite of smaller dicts in Python 3, we see how nicely __slots__ scale for instances to save us memory, and that is a major reason you would want to use __slots__.

Just for completeness of my notes, note that there is a one-time cost per slot in the class"s namespace of 64 bytes in Python 2, and 72 bytes in Python 3, because slots use data descriptors like properties, called "members".

>>> Foo.foo
<member "foo" of "Foo" objects>
>>> type(Foo.foo)
<class "member_descriptor">
>>> getsizeof(Foo.foo)
72

Demonstration of __slots__:

To deny the creation of a __dict__, you must subclass object. Everything subclasses object in Python 3, but in Python 2 you had to be explicit:

class Base(object): 
    __slots__ = ()

now:

>>> b = Base()
>>> b.a = "a"
Traceback (most recent call last):
  File "<pyshell#38>", line 1, in <module>
    b.a = "a"
AttributeError: "Base" object has no attribute "a"

Or subclass another class that defines __slots__

class Child(Base):
    __slots__ = ("a",)

and now:

c = Child()
c.a = "a"

but:

>>> c.b = "b"
Traceback (most recent call last):
  File "<pyshell#42>", line 1, in <module>
    c.b = "b"
AttributeError: "Child" object has no attribute "b"

To allow __dict__ creation while subclassing slotted objects, just add "__dict__" to the __slots__ (note that slots are ordered, and you shouldn"t repeat slots that are already in parent classes):

class SlottedWithDict(Child): 
    __slots__ = ("__dict__", "b")

swd = SlottedWithDict()
swd.a = "a"
swd.b = "b"
swd.c = "c"

and

>>> swd.__dict__
{"c": "c"}

Or you don"t even need to declare __slots__ in your subclass, and you will still use slots from the parents, but not restrict the creation of a __dict__:

class NoSlots(Child): pass
ns = NoSlots()
ns.a = "a"
ns.b = "b"

And:

>>> ns.__dict__
{"b": "b"}

However, __slots__ may cause problems for multiple inheritance:

class BaseA(object): 
    __slots__ = ("a",)

class BaseB(object): 
    __slots__ = ("b",)

Because creating a child class from parents with both non-empty slots fails:

>>> class Child(BaseA, BaseB): __slots__ = ()
Traceback (most recent call last):
  File "<pyshell#68>", line 1, in <module>
    class Child(BaseA, BaseB): __slots__ = ()
TypeError: Error when calling the metaclass bases
    multiple bases have instance lay-out conflict

If you run into this problem, You could just remove __slots__ from the parents, or if you have control of the parents, give them empty slots, or refactor to abstractions:

from abc import ABC

class AbstractA(ABC):
    __slots__ = ()

class BaseA(AbstractA): 
    __slots__ = ("a",)

class AbstractB(ABC):
    __slots__ = ()

class BaseB(AbstractB): 
    __slots__ = ("b",)

class Child(AbstractA, AbstractB): 
    __slots__ = ("a", "b")

c = Child() # no problem!

Add "__dict__" to __slots__ to get dynamic assignment:

class Foo(object):
    __slots__ = "bar", "baz", "__dict__"

and now:

>>> foo = Foo()
>>> foo.boink = "boink"

So with "__dict__" in slots we lose some of the size benefits with the upside of having dynamic assignment and still having slots for the names we do expect.

When you inherit from an object that isn"t slotted, you get the same sort of semantics when you use __slots__ - names that are in __slots__ point to slotted values, while any other values are put in the instance"s __dict__.

Avoiding __slots__ because you want to be able to add attributes on the fly is actually not a good reason - just add "__dict__" to your __slots__ if this is required.

You can similarly add __weakref__ to __slots__ explicitly if you need that feature.

Set to empty tuple when subclassing a namedtuple:

The namedtuple builtin make immutable instances that are very lightweight (essentially, the size of tuples) but to get the benefits, you need to do it yourself if you subclass them:

from collections import namedtuple
class MyNT(namedtuple("MyNT", "bar baz")):
    """MyNT is an immutable and lightweight object"""
    __slots__ = ()

usage:

>>> nt = MyNT("bar", "baz")
>>> nt.bar
"bar"
>>> nt.baz
"baz"

And trying to assign an unexpected attribute raises an AttributeError because we have prevented the creation of __dict__:

>>> nt.quux = "quux"
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: "MyNT" object has no attribute "quux"

You can allow __dict__ creation by leaving off __slots__ = (), but you can"t use non-empty __slots__ with subtypes of tuple.

Biggest Caveat: Multiple inheritance

Even when non-empty slots are the same for multiple parents, they cannot be used together:

class Foo(object): 
    __slots__ = "foo", "bar"
class Bar(object):
    __slots__ = "foo", "bar" # alas, would work if empty, i.e. ()

>>> class Baz(Foo, Bar): pass
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: Error when calling the metaclass bases
    multiple bases have instance lay-out conflict

Using an empty __slots__ in the parent seems to provide the most flexibility, allowing the child to choose to prevent or allow (by adding "__dict__" to get dynamic assignment, see section above) the creation of a __dict__:

class Foo(object): __slots__ = ()
class Bar(object): __slots__ = ()
class Baz(Foo, Bar): __slots__ = ("foo", "bar")
b = Baz()
b.foo, b.bar = "foo", "bar"

You don"t have to have slots - so if you add them, and remove them later, it shouldn"t cause any problems.

Going out on a limb here: If you"re composing mixins or using abstract base classes, which aren"t intended to be instantiated, an empty __slots__ in those parents seems to be the best way to go in terms of flexibility for subclassers.

To demonstrate, first, let"s create a class with code we"d like to use under multiple inheritance

class AbstractBase:
    __slots__ = ()
    def __init__(self, a, b):
        self.a = a
        self.b = b
    def __repr__(self):
        return f"{type(self).__name__}({repr(self.a)}, {repr(self.b)})"

We could use the above directly by inheriting and declaring the expected slots:

class Foo(AbstractBase):
    __slots__ = "a", "b"

But we don"t care about that, that"s trivial single inheritance, we need another class we might also inherit from, maybe with a noisy attribute:

class AbstractBaseC:
    __slots__ = ()
    @property
    def c(self):
        print("getting c!")
        return self._c
    @c.setter
    def c(self, arg):
        print("setting c!")
        self._c = arg

Now if both bases had nonempty slots, we couldn"t do the below. (In fact, if we wanted, we could have given AbstractBase nonempty slots a and b, and left them out of the below declaration - leaving them in would be wrong):

class Concretion(AbstractBase, AbstractBaseC):
    __slots__ = "a b _c".split()

And now we have functionality from both via multiple inheritance, and can still deny __dict__ and __weakref__ instantiation:

>>> c = Concretion("a", "b")
>>> c.c = c
setting c!
>>> c.c
getting c!
Concretion("a", "b")
>>> c.d = "d"
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: "Concretion" object has no attribute "d"

Other cases to avoid slots:

• Avoid them when you want to perform __class__ assignment with another class that doesn"t have them (and you can"t add them) unless the slot layouts are identical. (I am very interested in learning who is doing this and why.)
• Avoid them if you want to subclass variable length builtins like long, tuple, or str, and you want to add attributes to them.
• Avoid them if you insist on providing default values via class attributes for instance variables.

You may be able to tease out further caveats from the rest of the __slots__ documentation (the 3.7 dev docs are the most current), which I have made significant recent contributions to.

Critiques of other answers

The current top answers cite outdated information and are quite hand-wavy and miss the mark in some important ways.

Do not "only use __slots__ when instantiating lots of objects"

I quote:

"You would want to use __slots__ if you are going to instantiate a lot (hundreds, thousands) of objects of the same class."

Abstract Base Classes, for example, from the collections module, are not instantiated, yet __slots__ are declared for them.

Why?

If a user wishes to deny __dict__ or __weakref__ creation, those things must not be available in the parent classes.

__slots__ contributes to reusability when creating interfaces or mixins.

It is true that many Python users aren"t writing for reusability, but when you are, having the option to deny unnecessary space usage is valuable.

__slots__ doesn"t break pickling

When pickling a slotted object, you may find it complains with a misleading TypeError:

>>> pickle.loads(pickle.dumps(f))
TypeError: a class that defines __slots__ without defining __getstate__ cannot be pickled

This is actually incorrect. This message comes from the oldest protocol, which is the default. You can select the latest protocol with the -1 argument. In Python 2.7 this would be 2 (which was introduced in 2.3), and in 3.6 it is 4.

>>> pickle.loads(pickle.dumps(f, -1))
<__main__.Foo object at 0x1129C770>

in Python 2.7:

>>> pickle.loads(pickle.dumps(f, 2))
<__main__.Foo object at 0x1129C770>

in Python 3.6

>>> pickle.loads(pickle.dumps(f, 4))
<__main__.Foo object at 0x1129C770>

So I would keep this in mind, as it is a solved problem.

Critique of the (until Oct 2, 2016) accepted answer

The first paragraph is half short explanation, half predictive. Here"s the only part that actually answers the question

The proper use of __slots__ is to save space in objects. Instead of having a dynamic dict that allows adding attributes to objects at anytime, there is a static structure which does not allow additions after creation. This saves the overhead of one dict for every object that uses slots

The second half is wishful thinking, and off the mark:

While this is sometimes a useful optimization, it would be completely unnecessary if the Python interpreter was dynamic enough so that it would only require the dict when there actually were additions to the object.

Python actually does something similar to this, only creating the __dict__ when it is accessed, but creating lots of objects with no data is fairly ridiculous.

The second paragraph oversimplifies and misses actual reasons to avoid __slots__. The below is not a real reason to avoid slots (for actual reasons, see the rest of my answer above.):

They change the behavior of the objects that have slots in a way that can be abused by control freaks and static typing weenies.

It then goes on to discuss other ways of accomplishing that perverse goal with Python, not discussing anything to do with __slots__.

The third paragraph is more wishful thinking. Together it is mostly off-the-mark content that the answerer didn"t even author and contributes to ammunition for critics of the site.

Memory usage evidence

Create some normal objects and slotted objects:

>>> class Foo(object): pass
>>> class Bar(object): __slots__ = ()

Instantiate a million of them:

>>> foos = [Foo() for f in xrange(1000000)]
>>> bars = [Bar() for b in xrange(1000000)]

Inspect with guppy.hpy().heap():

>>> guppy.hpy().heap()
Partition of a set of 2028259 objects. Total size = 99763360 bytes.
 Index  Count   %     Size   % Cumulative  % Kind (class / dict of class)
     0 1000000  49 64000000  64  64000000  64 __main__.Foo
     1     169   0 16281480  16  80281480  80 list
     2 1000000  49 16000000  16  96281480  97 __main__.Bar
     3   12284   1   987472   1  97268952  97 str
...

Access the regular objects and their __dict__ and inspect again:

>>> for f in foos:
...     f.__dict__
>>> guppy.hpy().heap()
Partition of a set of 3028258 objects. Total size = 379763480 bytes.
 Index  Count   %      Size    % Cumulative  % Kind (class / dict of class)
     0 1000000  33 280000000  74 280000000  74 dict of __main__.Foo
     1 1000000  33  64000000  17 344000000  91 __main__.Foo
     2     169   0  16281480   4 360281480  95 list
     3 1000000  33  16000000   4 376281480  99 __main__.Bar
     4   12284   0    987472   0 377268952  99 str
...

This is consistent with the history of Python, from Unifying types and classes in Python 2.2

If you subclass a built-in type, extra space is automatically added to the instances to accomodate __dict__ and __weakrefs__. (The __dict__ is not initialized until you use it though, so you shouldn"t worry about the space occupied by an empty dictionary for each instance you create.) If you don"t need this extra space, you can add the phrase "__slots__ = []" to your class.

Answer #5

This is the behaviour to adopt when the referenced object is deleted. It is not specific to Django; this is an SQL standard. Although Django has its own implementation on top of SQL. (1)

There are seven possible actions to take when such event occurs:

• CASCADE: When the referenced object is deleted, also delete the objects that have references to it (when you remove a blog post for instance, you might want to delete comments as well). SQL equivalent: CASCADE.
• PROTECT: Forbid the deletion of the referenced object. To delete it you will have to delete all objects that reference it manually. SQL equivalent: RESTRICT.
• RESTRICT: (introduced in Django 3.1) Similar behavior as PROTECT that matches SQL"s RESTRICT more accurately. (See django documentation example)
• SET_NULL: Set the reference to NULL (requires the field to be nullable). For instance, when you delete a User, you might want to keep the comments he posted on blog posts, but say it was posted by an anonymous (or deleted) user. SQL equivalent: SET NULL.
• SET_DEFAULT: Set the default value. SQL equivalent: SET DEFAULT.
• SET(...): Set a given value. This one is not part of the SQL standard and is entirely handled by Django.
• DO_NOTHING: Probably a very bad idea since this would create integrity issues in your database (referencing an object that actually doesn"t exist). SQL equivalent: NO ACTION. (2)

Source: Django documentation

See also the documentation of PostgreSQL for instance.

In most cases, CASCADE is the expected behaviour, but for every ForeignKey, you should always ask yourself what is the expected behaviour in this situation. PROTECT and SET_NULL are often useful. Setting CASCADE where it should not, can potentially delete all of your database in cascade, by simply deleting a single user.

---

Additional note to clarify cascade direction

It"s funny to notice that the direction of the CASCADE action is not clear to many people. Actually, it"s funny to notice that only the CASCADE action is not clear. I understand the cascade behavior might be confusing, however you must think that it is the same direction as any other action. Thus, if you feel that CASCADE direction is not clear to you, it actually means that on_delete behavior is not clear to you.

In your database, a foreign key is basically represented by an integer field which value is the primary key of the foreign object. Let"s say you have an entry comment_A, which has a foreign key to an entry article_B. If you delete the entry comment_A, everything is fine. article_B used to live without comment_A and don"t bother if it"s deleted. However, if you delete article_B, then comment_A panics! It never lived without article_B and needs it, and it"s part of its attributes (article=article_B, but what is article_B???). This is where on_delete steps in, to determine how to resolve this integrity error, either by saying:

• "No! Please! Don"t! I can"t live without you!" (which is said PROTECT or RESTRICT in Django/SQL)
• "All right, if I"m not yours, then I"m nobody"s" (which is said SET_NULL)
• "Good bye world, I can"t live without article_B" and commit suicide (this is the CASCADE behavior).
• "It"s OK, I"ve got spare lover, and I"ll reference article_C from now" (SET_DEFAULT, or even SET(...)).
• "I can"t face reality, and I"ll keep calling your name even if that"s the only thing left to me!" (DO_NOTHING)

I hope it makes cascade direction clearer. :)

---

Footnotes

(1) Django has its own implementation on top of SQL. And, as mentioned by @JoeMjr2 in the comments below, Django will not create the SQL constraints. If you want the constraints to be ensured by your database (for instance, if your database is used by another application, or if you hang in the database console from time to time), you might want to set the related constraints manually yourself. There is an open ticket to add support for database-level on delete constrains in Django.

(2) Actually, there is one case where DO_NOTHING can be useful: If you want to skip Django"s implementation and implement the constraint yourself at the database-level.

Answer #6

TL;DR: If you are using Python 3.10 or later, it just works. As of today (2019), in 3.7+ you must turn this feature on using a future statement (from __future__ import annotations). In Python 3.6 or below, use a string.

I guess you got this exception:

NameError: name "Position" is not defined

This is because Position must be defined before you can use it in an annotation unless you are using Python 3.10 or later.

Python 3.7+: from __future__ import annotations

Python 3.7 introduces PEP 563: postponed evaluation of annotations. A module that uses the future statement from __future__ import annotations will store annotations as strings automatically:

from __future__ import annotations

class Position:
    def __add__(self, other: Position) -> Position:
        ...

This is scheduled to become the default in Python 3.10. Since Python still is a dynamically typed language so no type checking is done at runtime, typing annotations should have no performance impact, right? Wrong! Before python 3.7 the typing module used to be one of the slowest python modules in core so if you import typing you will see up to 7 times increase in performance when you upgrade to 3.7.

Python <3.7: use a string

According to PEP 484, you should use a string instead of the class itself:

class Position:
    ...
    def __add__(self, other: "Position") -> "Position":
       ...

If you use the Django framework this may be familiar as Django models also use strings for forward references (foreign key definitions where the foreign model is self or is not declared yet). This should work with Pycharm and other tools.

Sources

The relevant parts of PEP 484 and PEP 563, to spare you the trip:

Forward references

When a type hint contains names that have not been defined yet, that definition may be expressed as a string literal, to be resolved later.

A situation where this occurs commonly is the definition of a container class, where the class being defined occurs in the signature of some of the methods. For example, the following code (the start of a simple binary tree implementation) does not work:

class Tree:
    def __init__(self, left: Tree, right: Tree):
        self.left = left
        self.right = right

To address this, we write:

class Tree:
    def __init__(self, left: "Tree", right: "Tree"):
        self.left = left
        self.right = right

The string literal should contain a valid Python expression (i.e., compile(lit, "", "eval") should be a valid code object) and it should evaluate without errors once the module has been fully loaded. The local and global namespace in which it is evaluated should be the same namespaces in which default arguments to the same function would be evaluated.

and PEP 563:

Implementation

In Python 3.10, function and variable annotations will no longer be evaluated at definition time. Instead, a string form will be preserved in the respective¬†__annotations__¬†dictionary. Static type checkers will see no difference in behavior, whereas tools using annotations at runtime will have to perform postponed evaluation.

...

Enabling the future behavior in Python 3.7

The functionality described above can be enabled starting from Python 3.7 using the following special import:

from __future__ import annotations

Things that you may be tempted to do instead

A. Define a dummy Position

Before the class definition, place a dummy definition:

class Position(object):
    pass


class Position(object):
    ...

This will get rid of the NameError and may even look OK:

>>> Position.__add__.__annotations__
{"other": __main__.Position, "return": __main__.Position}

But is it?

>>> for k, v in Position.__add__.__annotations__.items():
...     print(k, "is Position:", v is Position)                                                                                                                                                                                                                  
return is Position: False
other is Position: False

B. Monkey-patch in order to add the annotations:

You may want to try some Python meta programming magic and write a decorator to monkey-patch the class definition in order to add annotations:

class Position:
    ...
    def __add__(self, other):
        return self.__class__(self.x + other.x, self.y + other.y)

The decorator should be responsible for the equivalent of this:

Position.__add__.__annotations__["return"] = Position
Position.__add__.__annotations__["other"] = Position

At least it seems right:

>>> for k, v in Position.__add__.__annotations__.items():
...     print(k, "is Position:", v is Position)                                                                                                                                                                                                                  
return is Position: True
other is Position: True

Probably too much trouble.

Answer #7

Is there any reason for a class declaration to inherit from object?

In Python 3, apart from compatibility between Python 2 and 3, no reason. In Python 2, many reasons.

---

Python 2.x story:

In Python 2.x (from 2.2 onwards) there"s two styles of classes depending on the presence or absence of object as a base-class:

1. "classic" style classes: they don"t have object as a base class:

   >>> class ClassicSpam:      # no base class
   ...     pass
   >>> ClassicSpam.__bases__
   ()
   

2. "new" style classes: they have, directly or indirectly (e.g inherit from a built-in type), object as a base class:

   >>> class NewSpam(object):           # directly inherit from object
   ...    pass
   >>> NewSpam.__bases__
   (<type "object">,)
   >>> class IntSpam(int):              # indirectly inherit from object...
   ...    pass
   >>> IntSpam.__bases__
   (<type "int">,) 
   >>> IntSpam.__bases__[0].__bases__   # ... because int inherits from object  
   (<type "object">,)
   

Without a doubt, when writing a class you"ll always want to go for new-style classes. The perks of doing so are numerous, to list some of them:

• Support for descriptors. Specifically, the following constructs are made possible with descriptors:

  1. classmethod: A method that receives the class as an implicit argument instead of the instance.
  2. staticmethod: A method that does not receive the implicit argument self as a first argument.
  3. properties with property: Create functions for managing the getting, setting and deleting of an attribute.
  4. __slots__: Saves memory consumptions of a class and also results in faster attribute access. Of course, it does impose limitations.

• The __new__ static method: lets you customize how new class instances are created.

• Method resolution order (MRO): in what order the base classes of a class will be searched when trying to resolve which method to call.

• Related to MRO, super calls. Also see, super() considered super.

If you don"t inherit from object, forget these. A more exhaustive description of the previous bullet points along with other perks of "new" style classes can be found here.

One of the downsides of new-style classes is that the class itself is more memory demanding. Unless you"re creating many class objects, though, I doubt this would be an issue and it"s a negative sinking in a sea of positives.

---

Python 3.x story:

In Python 3, things are simplified. Only new-style classes exist (referred to plainly as classes) so, the only difference in adding object is requiring you to type in 8 more characters. This:

class ClassicSpam:
    pass

is completely equivalent (apart from their name :-) to this:

class NewSpam(object):
     pass

and to this:

class Spam():
    pass

All have object in their __bases__.

>>> [object in cls.__bases__ for cls in {Spam, NewSpam, ClassicSpam}]
[True, True, True]

---

So, what should you do?

In Python 2: always inherit from object explicitly. Get the perks.

In Python 3: inherit from object if you are writing code that tries to be Python agnostic, that is, it needs to work both in Python 2 and in Python 3. Otherwise don"t, it really makes no difference since Python inserts it for you behind the scenes.

Answer #8

It must relate to the renaming and deprecation of cross_validation sub-module to model_selection. Try substituting cross_validation to model_selection

Answer #9

There are many ways to convert an instance to a dictionary, with varying degrees of corner case handling and closeness to the desired result.

---

1. instance.__dict__

instance.__dict__

which returns

{"_foreign_key_cache": <OtherModel: OtherModel object>,
 "_state": <django.db.models.base.ModelState at 0x7ff0993f6908>,
 "auto_now_add": datetime.datetime(2018, 12, 20, 21, 34, 29, 494827, tzinfo=<UTC>),
 "foreign_key_id": 2,
 "id": 1,
 "normal_value": 1,
 "readonly_value": 2}

This is by far the simplest, but is missing many_to_many, foreign_key is misnamed, and it has two unwanted extra things in it.

---

2. model_to_dict

from django.forms.models import model_to_dict
model_to_dict(instance)

which returns

{"foreign_key": 2,
 "id": 1,
 "many_to_many": [<OtherModel: OtherModel object>],
 "normal_value": 1}

This is the only one with many_to_many, but is missing the uneditable fields.

---

3. model_to_dict(..., fields=...)

from django.forms.models import model_to_dict
model_to_dict(instance, fields=[field.name for field in instance._meta.fields])

which returns

{"foreign_key": 2, "id": 1, "normal_value": 1}

This is strictly worse than the standard model_to_dict invocation.

---

4. query_set.values()

SomeModel.objects.filter(id=instance.id).values()[0]

which returns

{"auto_now_add": datetime.datetime(2018, 12, 20, 21, 34, 29, 494827, tzinfo=<UTC>),
 "foreign_key_id": 2,
 "id": 1,
 "normal_value": 1,
 "readonly_value": 2}

This is the same output as instance.__dict__ but without the extra fields. foreign_key_id is still wrong and many_to_many is still missing.

---

5. Custom Function

The code for django"s model_to_dict had most of the answer. It explicitly removed non-editable fields, so removing that check and getting the ids of foreign keys for many to many fields results in the following code which behaves as desired:

from itertools import chain

def to_dict(instance):
    opts = instance._meta
    data = {}
    for f in chain(opts.concrete_fields, opts.private_fields):
        data[f.name] = f.value_from_object(instance)
    for f in opts.many_to_many:
        data[f.name] = [i.id for i in f.value_from_object(instance)]
    return data

While this is the most complicated option, calling to_dict(instance) gives us exactly the desired result:

{"auto_now_add": datetime.datetime(2018, 12, 20, 21, 34, 29, 494827, tzinfo=<UTC>),
 "foreign_key": 2,
 "id": 1,
 "many_to_many": [2],
 "normal_value": 1,
 "readonly_value": 2}

---

6. Use Serializers

Django Rest Framework"s ModelSerialzer allows you to build a serializer automatically from a model.

from rest_framework import serializers
class SomeModelSerializer(serializers.ModelSerializer):
    class Meta:
        model = SomeModel
        fields = "__all__"

SomeModelSerializer(instance).data

returns

{"auto_now_add": "2018-12-20T21:34:29.494827Z",
 "foreign_key": 2,
 "id": 1,
 "many_to_many": [2],
 "normal_value": 1,
 "readonly_value": 2}

This is almost as good as the custom function, but auto_now_add is a string instead of a datetime object.

---

Bonus Round: better model printing

If you want a django model that has a better python command-line display, have your models child-class the following:

from django.db import models
from itertools import chain

class PrintableModel(models.Model):
    def __repr__(self):
        return str(self.to_dict())

    def to_dict(instance):
        opts = instance._meta
        data = {}
        for f in chain(opts.concrete_fields, opts.private_fields):
            data[f.name] = f.value_from_object(instance)
        for f in opts.many_to_many:
            data[f.name] = [i.id for i in f.value_from_object(instance)]
        return data

    class Meta:
        abstract = True

So, for example, if we define our models as such:

class OtherModel(PrintableModel): pass

class SomeModel(PrintableModel):
    normal_value = models.IntegerField()
    readonly_value = models.IntegerField(editable=False)
    auto_now_add = models.DateTimeField(auto_now_add=True)
    foreign_key = models.ForeignKey(OtherModel, related_name="ref1")
    many_to_many = models.ManyToManyField(OtherModel, related_name="ref2")

Calling SomeModel.objects.first() now gives output like this:

{"auto_now_add": datetime.datetime(2018, 12, 20, 21, 34, 29, 494827, tzinfo=<UTC>),
 "foreign_key": 2,
 "id": 1,
 "many_to_many": [2],
 "normal_value": 1,
 "readonly_value": 2}

Answer #10

The short answer, or TL;DR

Basically, eval is used to evaluate a single dynamically generated Python expression, and exec is used to execute dynamically generated Python code only for its side effects.

eval and exec have these two differences:

1. eval accepts only a single expression, exec can take a code block that has Python statements: loops, try: except:, class and function/method definitions and so on.

   An expression in Python is whatever you can have as the value in a variable assignment:

   a_variable = (anything you can put within these parentheses is an expression)
   

2. eval returns the value of the given expression, whereas exec ignores the return value from its code, and always returns None (in Python 2 it is a statement and cannot be used as an expression, so it really does not return anything).

In versions 1.0 - 2.7, exec was a statement, because CPython needed to produce a different kind of code object for functions that used exec for its side effects inside the function.

In Python 3, exec is a function; its use has no effect on the compiled bytecode of the function where it is used.

---

Thus basically:

>>> a = 5
>>> eval("37 + a")   # it is an expression
42
>>> exec("37 + a")   # it is an expression statement; value is ignored (None is returned)
>>> exec("a = 47")   # modify a global variable as a side effect
>>> a
47
>>> eval("a = 47")  # you cannot evaluate a statement
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 1
    a = 47
      ^
SyntaxError: invalid syntax

---

The compile in "exec" mode compiles any number of statements into a bytecode that implicitly always returns None, whereas in "eval" mode it compiles a single expression into bytecode that returns the value of that expression.

>>> eval(compile("42", "<string>", "exec"))  # code returns None
>>> eval(compile("42", "<string>", "eval"))  # code returns 42
42
>>> exec(compile("42", "<string>", "eval"))  # code returns 42,
>>>                                          # but ignored by exec

In the "eval" mode (and thus with the eval function if a string is passed in), the compile raises an exception if the source code contains statements or anything else beyond a single expression:

>>> compile("for i in range(3): print(i)", "<string>", "eval")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 1
    for i in range(3): print(i)
      ^
SyntaxError: invalid syntax

---

Actually the statement "eval accepts only a single expression" applies only when a string (which contains Python source code) is passed to eval. Then it is internally compiled to bytecode using compile(source, "<string>", "eval") This is where the difference really comes from.

If a code object (which contains Python bytecode) is passed to exec or eval, they behave identically, excepting for the fact that exec ignores the return value, still returning None always. So it is possible use eval to execute something that has statements, if you just compiled it into bytecode before instead of passing it as a string:

>>> eval(compile("if 1: print("Hello")", "<string>", "exec"))
Hello
>>>

works without problems, even though the compiled code contains statements. It still returns None, because that is the return value of the code object returned from compile.

In the "eval" mode (and thus with the eval function if a string is passed in), the compile raises an exception if the source code contains statements or anything else beyond a single expression:

>>> compile("for i in range(3): print(i)", "<string>". "eval")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 1
    for i in range(3): print(i)
      ^
SyntaxError: invalid syntax

The longer answer, a.k.a the gory details

exec and eval

The exec function (which was a statement in Python 2) is used for executing a dynamically created statement or program:

>>> program = """
for i in range(3):
    print("Python is cool")
"""
>>> exec(program)
Python is cool
Python is cool
Python is cool
>>> 

The eval function does the same for a single expression, and returns the value of the expression:

>>> a = 2
>>> my_calculation = "42 * a"
>>> result = eval(my_calculation)
>>> result
84

exec and eval both accept the program/expression to be run either as a str, unicode or bytes object containing source code, or as a code object which contains Python bytecode.

If a str/unicode/bytes containing source code was passed to exec, it behaves equivalently to:

exec(compile(source, "<string>", "exec"))

and eval similarly behaves equivalent to:

eval(compile(source, "<string>", "eval"))

---

Since all expressions can be used as statements in Python (these are called the Expr nodes in the Python abstract grammar; the opposite is not true), you can always use exec if you do not need the return value. That is to say, you can use either eval("my_func(42)") or exec("my_func(42)"), the difference being that eval returns the value returned by my_func, and exec discards it:

>>> def my_func(arg):
...     print("Called with %d" % arg)
...     return arg * 2
... 
>>> exec("my_func(42)")
Called with 42
>>> eval("my_func(42)")
Called with 42
84
>>> 

Of the 2, only exec accepts source code that contains statements, like def, for, while, import, or class, the assignment statement (a.k.a a = 42), or entire programs:

>>> exec("for i in range(3): print(i)")
0
1
2
>>> eval("for i in range(3): print(i)")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 1
    for i in range(3): print(i)
      ^
SyntaxError: invalid syntax

---

Both exec and eval accept 2 additional positional arguments - globals and locals - which are the global and local variable scopes that the code sees. These default to the globals() and locals() within the scope that called exec or eval, but any dictionary can be used for globals and any mapping for locals (including dict of course). These can be used not only to restrict/modify the variables that the code sees, but are often also used for capturing the variables that the executed code creates:

>>> g = dict()
>>> l = dict()
>>> exec("global a; a, b = 123, 42", g, l)
>>> g["a"]
123
>>> l
{"b": 42}

(If you display the value of the entire g, it would be much longer, because exec and eval add the built-ins module as __builtins__ to the globals automatically if it is missing).

In Python 2, the official syntax for the exec statement is actually exec code in globals, locals, as in

>>> exec "global a; a, b = 123, 42" in g, l

However the alternate syntax exec(code, globals, locals) has always been accepted too (see below).

compile

The compile(source, filename, mode, flags=0, dont_inherit=False, optimize=-1) built-in can be used to speed up repeated invocations of the same code with exec or eval by compiling the source into a code object beforehand. The mode parameter controls the kind of code fragment the compile function accepts and the kind of bytecode it produces. The choices are "eval", "exec" and "single":

• "eval" mode expects a single expression, and will produce bytecode that when run will return the value of that expression:

  >>> dis.dis(compile("a + b", "<string>", "eval"))
    1           0 LOAD_NAME                0 (a)
                3 LOAD_NAME                1 (b)
                6 BINARY_ADD
                7 RETURN_VALUE
  

• "exec" accepts any kinds of python constructs from single expressions to whole modules of code, and executes them as if they were module top-level statements. The code object returns None:

  >>> dis.dis(compile("a + b", "<string>", "exec"))
    1           0 LOAD_NAME                0 (a)
                3 LOAD_NAME                1 (b)
                6 BINARY_ADD
                7 POP_TOP                             <- discard result
                8 LOAD_CONST               0 (None)   <- load None on stack
               11 RETURN_VALUE                        <- return top of stack
  

• "single" is a limited form of "exec" which accepts a source code containing a single statement (or multiple statements separated by ;) if the last statement is an expression statement, the resulting bytecode also prints the repr of the value of that expression to the standard output(!).

  An if-elif-else chain, a loop with else, and try with its except, else and finally blocks is considered a single statement.

  A source fragment containing 2 top-level statements is an error for the "single", except in Python 2 there is a bug that sometimes allows multiple toplevel statements in the code; only the first is compiled; the rest are ignored:

  In Python 2.7.8:

  >>> exec(compile("a = 5
  a = 6", "<string>", "single"))
  >>> a
  5
  

  And in Python 3.4.2:

  >>> exec(compile("a = 5
  a = 6", "<string>", "single"))
  Traceback (most recent call last):
    File "<stdin>", line 1, in <module>
    File "<string>", line 1
      a = 5
          ^
  SyntaxError: multiple statements found while compiling a single statement
  

  This is very useful for making interactive Python shells. However, the value of the expression is not returned, even if you eval the resulting code.

Thus greatest distinction of exec and eval actually comes from the compile function and its modes.

---

In addition to compiling source code to bytecode, compile supports compiling abstract syntax trees (parse trees of Python code) into code objects; and source code into abstract syntax trees (the ast.parse is written in Python and just calls compile(source, filename, mode, PyCF_ONLY_AST)); these are used for example for modifying source code on the fly, and also for dynamic code creation, as it is often easier to handle the code as a tree of nodes instead of lines of text in complex cases.

---

While eval only allows you to evaluate a string that contains a single expression, you can eval a whole statement, or even a whole module that has been compiled into bytecode; that is, with Python 2, print is a statement, and cannot be evalled directly:

>>> eval("for i in range(3): print("Python is cool")")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 1
    for i in range(3): print("Python is cool")
      ^
SyntaxError: invalid syntax

compile it with "exec" mode into a code object and you can eval it; the eval function will return None.

>>> code = compile("for i in range(3): print("Python is cool")",
                   "foo.py", "exec")
>>> eval(code)
Python is cool
Python is cool
Python is cool

If one looks into eval and exec source code in CPython 3, this is very evident; they both call PyEval_EvalCode with same arguments, the only difference being that exec explicitly returns None.

Syntax differences of exec between Python 2 and Python 3

One of the major differences in Python 2 is that exec is a statement and eval is a built-in function (both are built-in functions in Python 3). It is a well-known fact that the official syntax of exec in Python 2 is exec code [in globals[, locals]].

Unlike majority of the Python 2-to-3 porting guides seem to suggest, the exec statement in CPython 2 can be also used with syntax that looks exactly like the exec function invocation in Python 3. The reason is that Python 0.9.9 had the exec(code, globals, locals) built-in function! And that built-in function was replaced with exec statement somewhere before Python 1.0 release.

Since it was desirable to not break backwards compatibility with Python 0.9.9, Guido van Rossum added a compatibility hack in 1993: if the code was a tuple of length 2 or 3, and globals and locals were not passed into the exec statement otherwise, the code would be interpreted as if the 2nd and 3rd element of the tuple were the globals and locals respectively. The compatibility hack was not mentioned even in Python 1.4 documentation (the earliest available version online); and thus was not known to many writers of the porting guides and tools, until it was documented again in November 2012:

The first expression may also be a tuple of length 2 or 3. In this case, the optional parts must be omitted. The form exec(expr, globals) is equivalent to exec expr in globals, while the form exec(expr, globals, locals) is equivalent to exec expr in globals, locals. The tuple form of exec provides compatibility with Python 3, where exec is a function rather than a statement.

Yes, in CPython 2.7 that it is handily referred to as being a forward-compatibility option (why confuse people over that there is a backward compatibility option at all), when it actually had been there for backward-compatibility for two decades.

Thus while exec is a statement in Python 1 and Python 2, and a built-in function in Python 3 and Python 0.9.9,

>>> exec("print(a)", globals(), {"a": 42})
42

has had identical behaviour in possibly every widely released Python version ever; and works in Jython 2.5.2, PyPy 2.3.1 (Python 2.7.6) and IronPython 2.6.1 too (kudos to them following the undocumented behaviour of CPython closely).

What you cannot do in Pythons 1.0 - 2.7 with its compatibility hack, is to store the return value of exec into a variable:

Python 2.7.11+ (default, Apr 17 2016, 14:00:29) 
[GCC 5.3.1 20160413] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> a = exec("print(42)")
  File "<stdin>", line 1
    a = exec("print(42)")
           ^
SyntaxError: invalid syntax

(which wouldn"t be useful in Python 3 either, as exec always returns None), or pass a reference to exec:

>>> call_later(exec, "print(42)", delay=1000)
  File "<stdin>", line 1
    call_later(exec, "print(42)", delay=1000)
                  ^
SyntaxError: invalid syntax

Which a pattern that someone might actually have used, though unlikely;

Or use it in a list comprehension:

>>> [exec(i) for i in ["print(42)", "print(foo)"]
  File "<stdin>", line 1
    [exec(i) for i in ["print(42)", "print(foo)"]
        ^
SyntaxError: invalid syntax

which is abuse of list comprehensions (use a for loop instead!).