Python Profilers 分析器?

源代碼： Lib/profile.py 和 Lib/pstats.py

profile分析器簡(jiǎn)介?

cProfile 和 profile 提供了 Python 程序的 確定性性能分析 。 profile 是一組統(tǒng)計(jì)數(shù)據(jù)，描述程序的各個(gè)部分執(zhí)行的頻率和時(shí)間。這些統(tǒng)計(jì)數(shù)據(jù)可以通過(guò) pstats 模塊格式化為報(bào)表。

Python 標(biāo)準(zhǔn)庫(kù)提供了同一分析接口的兩種不同實(shí)現(xiàn)：

  1. 對(duì)于大多數(shù)用戶，建議使用 cProfile ；這是一個(gè) C 擴(kuò)展插件，因?yàn)槠浜侠淼倪\(yùn)行開(kāi)銷，所以適合于分析長(zhǎng)時(shí)間運(yùn)行的程序。該插件基于 lsprof ，由 Brett Rosen 和 Ted Chaotter 貢獻(xiàn)。

  2. profile 是一個(gè)純 Python 模塊（cProfile 就是模擬其接口的 C 語(yǔ)言實(shí)現(xiàn)），但它會(huì)顯著增加配置程序的開(kāi)銷。如果你正在嘗試以某種方式擴(kuò)展分析器，則使用此模塊可能會(huì)更容易完成任務(wù)。該模塊最初由 Jim Roskind 設(shè)計(jì)和編寫。

備注

profiler 分析器模塊被設(shè)計(jì)為給指定的程序提供執(zhí)行概要文件，而不是用于基準(zhǔn)測(cè)試目的（ timeit 才是用于此目標(biāo)的，它能獲得合理準(zhǔn)確的結(jié)果）。這特別適用于將 Python 代碼與 C 代碼進(jìn)行基準(zhǔn)測(cè)試：分析器為Python 代碼引入開(kāi)銷，但不會(huì)為 C級(jí)別的函數(shù)引入開(kāi)銷，因此 C 代碼似乎比任何Python 代碼都更快。

實(shí)時(shí)用戶手冊(cè)?

本節(jié)是為 “不想閱讀手冊(cè)” 的用戶提供的。它提供了非常簡(jiǎn)短的概述，并允許用戶快速對(duì)現(xiàn)有應(yīng)用程序執(zhí)行評(píng)測(cè)。

要分析采用單個(gè)參數(shù)的函數(shù)，可以執(zhí)行以下操作：

import cProfile
import re
cProfile.run('re.compile("foo|bar")')

（如果 cProfile 在您的系統(tǒng)上不可用，請(qǐng)使用 profile 。）

上述操作將運(yùn)行 re.compile() 并打印分析結(jié)果，如下所示：

      214 function calls (207 primitive calls) in 0.002 seconds

Ordered by: cumulative time

ncalls  tottime  percall  cumtime  percall filename:lineno(function)
     1    0.000    0.000    0.002    0.002 {built-in method builtins.exec}
     1    0.000    0.000    0.001    0.001 <string>:1(<module>)
     1    0.000    0.000    0.001    0.001 __init__.py:250(compile)
     1    0.000    0.000    0.001    0.001 __init__.py:289(_compile)
     1    0.000    0.000    0.000    0.000 _compiler.py:759(compile)
     1    0.000    0.000    0.000    0.000 _parser.py:937(parse)
     1    0.000    0.000    0.000    0.000 _compiler.py:598(_code)
     1    0.000    0.000    0.000    0.000 _parser.py:435(_parse_sub)

The first line indicates that 214 calls were monitored. Of those calls, 207 were primitive, meaning that the call was not induced via recursion. The next line: Ordered by: cumulative name, indicates that the text string in the far right column was used to sort the output. The column headings include:

ncalls

調(diào)用次數(shù)

tottime

在指定函數(shù)中消耗的總時(shí)間（不包括調(diào)用子函數(shù)的時(shí)間）

percall

是 tottime 除以 ncalls 的商

cumtime

指定的函數(shù)及其所有子函數(shù)（從調(diào)用到退出）消耗的累積時(shí)間。這個(gè)數(shù)字對(duì)于遞歸函數(shù)來(lái)說(shuō)是準(zhǔn)確的。

percall

是 cumtime 除以原始調(diào)用（次數(shù)）的商（即：函數(shù)運(yùn)行一次的平均時(shí)間）

filename:lineno(function)

提供相應(yīng)數(shù)據(jù)的每個(gè)函數(shù)

如果第一列中有兩個(gè)數(shù)字（例如3/1），則表示函數(shù)遞歸。第二個(gè)值是原始調(diào)用次數(shù)，第一個(gè)是調(diào)用的總次數(shù)。請(qǐng)注意，當(dāng)函數(shù)不遞歸時(shí)，這兩個(gè)值是相同的，并且只打印單個(gè)數(shù)字。

profile 運(yùn)行結(jié)束時(shí)，打印輸出不是必須的。也可以通過(guò)為 run() 函數(shù)指定文件名，將結(jié)果保存到文件中：

import cProfile
import re
cProfile.run('re.compile("foo|bar")', 'restats')

pstats.Stats 類從文件中讀取 profile 結(jié)果，并以各種方式對(duì)其進(jìn)行格式化。

cProfile 和 profile 文件也可以作為腳本調(diào)用，以分析另一個(gè)腳本。例如：

python -m cProfile [-o output_file] [-s sort_order] (-m module | myscript.py)

-o 將profile 結(jié)果寫入文件而不是標(biāo)準(zhǔn)輸出

-s 指定 sort_stats() 排序值之一以對(duì)輸出進(jìn)行排序。這僅適用于未提供 -o 的情況

-m 指定要分析的是模塊而不是腳本。

3.7 新版功能: cProfile 添加 -m 選項(xiàng)

3.8 新版功能: profile 添加 -m 選項(xiàng)

The pstats module's Stats class has a variety of methods for manipulating and printing the data saved into a profile results file:

import pstats
from pstats import SortKey
p = pstats.Stats('restats')
p.strip_dirs().sort_stats(-1).print_stats()

The strip_dirs() method removed the extraneous path from all the module names. The sort_stats() method sorted all the entries according to the standard module/line/name string that is printed. The print_stats() method printed out all the statistics. You might try the following sort calls:

p.sort_stats(SortKey.NAME)
p.print_stats()

The first call will actually sort the list by function name, and the second call will print out the statistics. The following are some interesting calls to experiment with:

p.sort_stats(SortKey.CUMULATIVE).print_stats(10)

This sorts the profile by cumulative time in a function, and then only prints the ten most significant lines. If you want to understand what algorithms are taking time, the above line is what you would use.

If you were looking to see what functions were looping a lot, and taking a lot of time, you would do:

p.sort_stats(SortKey.TIME).print_stats(10)

to sort according to time spent within each function, and then print the statistics for the top ten functions.

你也可以嘗試：

p.sort_stats(SortKey.FILENAME).print_stats('__init__')

This will sort all the statistics by file name, and then print out statistics for only the class init methods (since they are spelled with __init__ in them). As one final example, you could try:

p.sort_stats(SortKey.TIME, SortKey.CUMULATIVE).print_stats(.5, 'init')

This line sorts statistics with a primary key of time, and a secondary key of cumulative time, and then prints out some of the statistics. To be specific, the list is first culled down to 50% (re: .5) of its original size, then only lines containing init are maintained, and that sub-sub-list is printed.

If you wondered what functions called the above functions, you could now (p is still sorted according to the last criteria) do:

p.print_callers(.5, 'init')

and you would get a list of callers for each of the listed functions.

If you want more functionality, you're going to have to read the manual, or guess what the following functions do:

p.print_callees()
p.add('restats')

Invoked as a script, the pstats module is a statistics browser for reading and examining profile dumps. It has a simple line-oriented interface (implemented using cmd) and interactive help.

profile 和 cProfile 模塊參考?

profile 和 cProfile 模塊都提供下列函數(shù)：

profile.run(command, filename=None, sort=- 1)?

This function takes a single argument that can be passed to the exec() function, and an optional file name. In all cases this routine executes:

exec(command, __main__.__dict__, __main__.__dict__)

and gathers profiling statistics from the execution. If no file name is present, then this function automatically creates a Stats instance and prints a simple profiling report. If the sort value is specified, it is passed to this Stats instance to control how the results are sorted.

profile.runctx(command, globals, locals, filename=None, sort=- 1)?

This function is similar to run(), with added arguments to supply the globals and locals dictionaries for the command string. This routine executes:

exec(command, globals, locals)

and gathers profiling statistics as in the run() function above.

class profile.Profile(timer=None, timeunit=0.0, subcalls=True, builtins=True)?

This class is normally only used if more precise control over profiling is needed than what the cProfile.run() function provides.

A custom timer can be supplied for measuring how long code takes to run via the timer argument. This must be a function that returns a single number representing the current time. If the number is an integer, the timeunit specifies a multiplier that specifies the duration of each unit of time. For example, if the timer returns times measured in thousands of seconds, the time unit would be .001.

Directly using the Profile class allows formatting profile results without writing the profile data to a file:

import cProfile, pstats, io
from pstats import SortKey
pr = cProfile.Profile()
pr.enable()
# ... do something ...
pr.disable()
s = io.StringIO()
sortby = SortKey.CUMULATIVE
ps = pstats.Stats(pr, stream=s).sort_stats(sortby)
ps.print_stats()
print(s.getvalue())

The Profile class can also be used as a context manager (supported only in cProfile module. see 上下文管理器類型):

import cProfile

with cProfile.Profile() as pr:
    # ... do something ...

pr.print_stats()

在 3.8 版更改: 添加了上下文管理器支持。

enable()?

開(kāi)始收集分析數(shù)據(jù)。僅在 cProfile 可用。

disable()?

停止收集分析數(shù)據(jù)。僅在 cProfile 可用。

create_stats()?

停止收集分析數(shù)據(jù)，并在內(nèi)部將結(jié)果記錄為當(dāng)前 profile。

print_stats(sort=- 1)?

Create a Stats object based on the current profile and print the results to stdout.

dump_stats(filename)?

將當(dāng)前profile 的結(jié)果寫入 filename 。

run(cmd)?

Profile the cmd via exec().

runctx(cmd, globals, locals)?

Profile the cmd via exec() with the specified global and local environment.

runcall(func, /, *args, **kwargs)?

Profile func(*args, **kwargs)

Note that profiling will only work if the called command/function actually returns. If the interpreter is terminated (e.g. via a sys.exit() call during the called command/function execution) no profiling results will be printed.

Stats 類?

Analysis of the profiler data is done using the Stats class.

class pstats.Stats(*filenames or profile, stream=sys.stdout)?

This class constructor creates an instance of a "statistics object" from a filename (or list of filenames) or from a Profile instance. Output will be printed to the stream specified by stream.

The file selected by the above constructor must have been created by the corresponding version of profile or cProfile. To be specific, there is no file compatibility guaranteed with future versions of this profiler, and there is no compatibility with files produced by other profilers, or the same profiler run on a different operating system. If several files are provided, all the statistics for identical functions will be coalesced, so that an overall view of several processes can be considered in a single report. If additional files need to be combined with data in an existing Stats object, the add() method can be used.

Instead of reading the profile data from a file, a cProfile.Profile or profile.Profile object can be used as the profile data source.

Stats 對(duì)象有以下方法:

strip_dirs()?

This method for the Stats class removes all leading path information from file names. It is very useful in reducing the size of the printout to fit within (close to) 80 columns. This method modifies the object, and the stripped information is lost. After performing a strip operation, the object is considered to have its entries in a "random" order, as it was just after object initialization and loading. If strip_dirs() causes two function names to be indistinguishable (they are on the same line of the same filename, and have the same function name), then the statistics for these two entries are accumulated into a single entry.

add(*filenames)?

This method of the Stats class accumulates additional profiling information into the current profiling object. Its arguments should refer to filenames created by the corresponding version of profile.run() or cProfile.run(). Statistics for identically named (re: file, line, name) functions are automatically accumulated into single function statistics.

dump_stats(filename)?

Save the data loaded into the Stats object to a file named filename. The file is created if it does not exist, and is overwritten if it already exists. This is equivalent to the method of the same name on the profile.Profile and cProfile.Profile classes.

sort_stats(*keys)?

This method modifies the Stats object by sorting it according to the supplied criteria. The argument can be either a string or a SortKey enum identifying the basis of a sort (example: 'time', 'name', SortKey.TIME or SortKey.NAME). The SortKey enums argument have advantage over the string argument in that it is more robust and less error prone.

When more than one key is provided, then additional keys are used as secondary criteria when there is equality in all keys selected before them. For example, sort_stats(SortKey.NAME, SortKey.FILE) will sort all the entries according to their function name, and resolve all ties (identical function names) by sorting by file name.

For the string argument, abbreviations can be used for any key names, as long as the abbreviation is unambiguous.

The following are the valid string and SortKey:

有效字符串參數(shù)

有效枚舉參數(shù)

含意

'calls'

SortKey.CALLS

調(diào)用次數(shù)

'cumulative'

SortKey.CUMULATIVE

累積時(shí)間

'cumtime'

N/A

累積時(shí)間

'file'

N/A

文件名

'filename'

SortKey.FILENAME

文件名

'module'

N/A

文件名

'ncalls'

N/A

調(diào)用次數(shù)

'pcalls'

SortKey.PCALLS

原始調(diào)用計(jì)數(shù)

'line'

SortKey.LINE

行號(hào)

'name'

SortKey.NAME

函數(shù)名稱

'nfl'

SortKey.NFL

名稱/文件/行

'stdname'

SortKey.STDNAME

標(biāo)準(zhǔn)名稱

'time'

SortKey.TIME

內(nèi)部時(shí)間

'tottime'

N/A

內(nèi)部時(shí)間

Note that all sorts on statistics are in descending order (placing most time consuming items first), where as name, file, and line number searches are in ascending order (alphabetical). The subtle distinction between SortKey.NFL and SortKey.STDNAME is that the standard name is a sort of the name as printed, which means that the embedded line numbers get compared in an odd way. For example, lines 3, 20, and 40 would (if the file names were the same) appear in the string order 20, 3 and 40. In contrast, SortKey.NFL does a numeric compare of the line numbers. In fact, sort_stats(SortKey.NFL) is the same as sort_stats(SortKey.NAME, SortKey.FILENAME, SortKey.LINE).

For backward-compatibility reasons, the numeric arguments -1, 0, 1, and 2 are permitted. They are interpreted as 'stdname', 'calls', 'time', and 'cumulative' respectively. If this old style format (numeric) is used, only one sort key (the numeric key) will be used, and additional arguments will be silently ignored.

3.7 新版功能: Added the SortKey enum.

reverse_order()?

This method for the Stats class reverses the ordering of the basic list within the object. Note that by default ascending vs descending order is properly selected based on the sort key of choice.

print_stats(*restrictions)?

This method for the Stats class prints out a report as described in the profile.run() definition.

The order of the printing is based on the last sort_stats() operation done on the object (subject to caveats in add() and strip_dirs()).

The arguments provided (if any) can be used to limit the list down to the significant entries. Initially, the list is taken to be the complete set of profiled functions. Each restriction is either an integer (to select a count of lines), or a decimal fraction between 0.0 and 1.0 inclusive (to select a percentage of lines), or a string that will interpreted as a regular expression (to pattern match the standard name that is printed). If several restrictions are provided, then they are applied sequentially. For example:

print_stats(.1, 'foo:')

would first limit the printing to first 10% of list, and then only print functions that were part of filename .*foo:. In contrast, the command:

print_stats('foo:', .1)

would limit the list to all functions having file names .*foo:, and then proceed to only print the first 10% of them.

print_callers(*restrictions)?

This method for the Stats class prints a list of all functions that called each function in the profiled database. The ordering is identical to that provided by print_stats(), and the definition of the restricting argument is also identical. Each caller is reported on its own line. The format differs slightly depending on the profiler that produced the stats:

  • With profile, a number is shown in parentheses after each caller to show how many times this specific call was made. For convenience, a second non-parenthesized number repeats the cumulative time spent in the function at the right.

  • With cProfile, each caller is preceded by three numbers: the number of times this specific call was made, and the total and cumulative times spent in the current function while it was invoked by this specific caller.

print_callees(*restrictions)?

This method for the Stats class prints a list of all function that were called by the indicated function. Aside from this reversal of direction of calls (re: called vs was called by), the arguments and ordering are identical to the print_callers() method.

get_stats_profile()?

This method returns an instance of StatsProfile, which contains a mapping of function names to instances of FunctionProfile. Each FunctionProfile instance holds information related to the function's profile such as how long the function took to run, how many times it was called, etc...

3.9 新版功能: Added the following dataclasses: StatsProfile, FunctionProfile. Added the following function: get_stats_profile.

什么是確定性性能分析？?

確定性性能分析 旨在反映這樣一個(gè)事實(shí)：即所有 函數(shù)調(diào)用 、 函數(shù)返回 和 異常 事件都被監(jiān)控，并且對(duì)這些事件之間的間隔（在此期間用戶的代碼正在執(zhí)行）進(jìn)行精確計(jì)時(shí)。相反，統(tǒng)計(jì)分析（不是由該模塊完成）隨機(jī)采樣有效指令指針，并推斷時(shí)間花費(fèi)在哪里。后一種技術(shù)傳統(tǒng)上涉及較少的開(kāi)銷（因?yàn)榇a不需要檢測(cè)），但只提供了時(shí)間花在哪里的相對(duì)指示。

在Python中，由于在執(zhí)行過(guò)程中總有一個(gè)活動(dòng)的解釋器，因此執(zhí)行確定性評(píng)測(cè)不需要插入指令的代碼。Python 自動(dòng)為每個(gè)事件提供一個(gè):dfn:鉤子 （可選回調(diào)）。此外，Python 的解釋特性往往會(huì)給執(zhí)行增加太多開(kāi)銷，以至于在典型的應(yīng)用程序中，確定性分析往往只會(huì)增加很小的處理開(kāi)銷。結(jié)果是，確定性分析并沒(méi)有那么代價(jià)高昂，但是它提供了有關(guān) Python 程序執(zhí)行的大量運(yùn)行時(shí)統(tǒng)計(jì)信息。

調(diào)用計(jì)數(shù)統(tǒng)計(jì)信息可用于識(shí)別代碼中的錯(cuò)誤（意外計(jì)數(shù)），并識(shí)別可能的內(nèi)聯(lián)擴(kuò)展點(diǎn)（高頻調(diào)用）。內(nèi)部時(shí)間統(tǒng)計(jì)可用于識(shí)別應(yīng)仔細(xì)優(yōu)化的 "熱循環(huán)" 。累積時(shí)間統(tǒng)計(jì)可用于識(shí)別算法選擇上的高級(jí)別錯(cuò)誤。請(qǐng)注意，該分析器中對(duì)累積時(shí)間的異常處理，允許直接比較算法的遞歸實(shí)現(xiàn)與迭代實(shí)現(xiàn)的統(tǒng)計(jì)信息。

局限性?

一個(gè)限制是關(guān)于時(shí)間信息的準(zhǔn)確性。確定性性能分析存在一個(gè)涉及精度的基本問(wèn)題。最明顯的限制是，底層的 "時(shí)鐘" 周期大約為0.001秒（通常）。因此，沒(méi)有什么測(cè)量會(huì)比底層時(shí)鐘更精確。如果進(jìn)行了足夠的測(cè)量，那么 "誤差" 將趨于平均。不幸的是，刪除第一個(gè)錯(cuò)誤會(huì)導(dǎo)致第二個(gè)錯(cuò)誤來(lái)源。

第二個(gè)問(wèn)題是，從調(diào)度事件到分析器調(diào)用獲取時(shí)間函數(shù)實(shí)際 獲取 時(shí)鐘狀態(tài)，這需要 "一段時(shí)間" 。類似地，從獲取時(shí)鐘值（然后保存）開(kāi)始，直到再次執(zhí)行用戶代碼為止，退出分析器事件句柄時(shí)也存在一定的延遲。因此，多次調(diào)用單個(gè)函數(shù)或調(diào)用多個(gè)函數(shù)通常會(huì)累積此錯(cuò)誤。盡管這種方式的誤差通常小于時(shí)鐘的精度（小于一個(gè)時(shí)鐘周期），但它 可以 累積并變得非?？捎^。

與開(kāi)銷較低的 cProfile 相比， profile 的問(wèn)題更為嚴(yán)重。出于這個(gè)原因， profile 提供了一種針對(duì)指定平臺(tái)的自我校準(zhǔn)方法，以便可以在很大程度上（平均地）消除此誤差。 校準(zhǔn)后，結(jié)果將更準(zhǔn)確（在最小二乘意義上），但它有時(shí)會(huì)產(chǎn)生負(fù)數(shù)（當(dāng)調(diào)用計(jì)數(shù)異常低，且概率之神對(duì)您不利時(shí)：-）。因此 不要 對(duì)產(chǎn)生的負(fù)數(shù)感到驚慌。它們應(yīng)該只在你手工校準(zhǔn)分析器的情況下才會(huì)出現(xiàn)，實(shí)際上結(jié)果比沒(méi)有校準(zhǔn)的情況要好。

準(zhǔn)確估量?

profile 模塊的 profiler 會(huì)從每個(gè)事件處理時(shí)間中減去一個(gè)常量，以補(bǔ)償調(diào)用 time 函數(shù)和存儲(chǔ)結(jié)果的開(kāi)銷。默認(rèn)情況下，常數(shù)為0。對(duì)于特定的平臺(tái)，可用以下程序獲得更好修正常數(shù)（ 局限性 ）。

import profile
pr = profile.Profile()
for i in range(5):
    print(pr.calibrate(10000))

The method executes the number of Python calls given by the argument, directly and again under the profiler, measuring the time for both. It then computes the hidden overhead per profiler event, and returns that as a float. For example, on a 1.8Ghz Intel Core i5 running macOS, and using Python's time.process_time() as the timer, the magical number is about 4.04e-6.

The object of this exercise is to get a fairly consistent result. If your computer is very fast, or your timer function has poor resolution, you might have to pass 100000, or even 1000000, to get consistent results.

當(dāng)你有一個(gè)一致的答案時(shí)，有三種方法可以使用：

import profile

# 1. Apply computed bias to all Profile instances created hereafter.
profile.Profile.bias = your_computed_bias

# 2. Apply computed bias to a specific Profile instance.
pr = profile.Profile()
pr.bias = your_computed_bias

# 3. Specify computed bias in instance constructor.
pr = profile.Profile(bias=your_computed_bias)

If you have a choice, you are better off choosing a smaller constant, and then your results will "less often" show up as negative in profile statistics.

使用自定義計(jì)時(shí)器?

If you want to change how current time is determined (for example, to force use of wall-clock time or elapsed process time), pass the timing function you want to the Profile class constructor:

pr = profile.Profile(your_time_func)

The resulting profiler will then call your_time_func. Depending on whether you are using profile.Profile or cProfile.Profile, your_time_func's return value will be interpreted differently:

profile.Profile

your_time_func should return a single number, or a list of numbers whose sum is the current time (like what os.times() returns). If the function returns a single time number, or the list of returned numbers has length 2, then you will get an especially fast version of the dispatch routine.

Be warned that you should calibrate the profiler class for the timer function that you choose (see 準(zhǔn)確估量). For most machines, a timer that returns a lone integer value will provide the best results in terms of low overhead during profiling. (os.times() is pretty bad, as it returns a tuple of floating point values). If you want to substitute a better timer in the cleanest fashion, derive a class and hardwire a replacement dispatch method that best handles your timer call, along with the appropriate calibration constant.

cProfile.Profile

your_time_func should return a single number. If it returns integers, you can also invoke the class constructor with a second argument specifying the real duration of one unit of time. For example, if your_integer_time_func returns times measured in thousands of seconds, you would construct the Profile instance as follows:

pr = cProfile.Profile(your_integer_time_func, 0.001)

As the cProfile.Profile class cannot be calibrated, custom timer functions should be used with care and should be as fast as possible. For the best results with a custom timer, it might be necessary to hard-code it in the C source of the internal _lsprof module.

Python 3.3 adds several new functions in time that can be used to make precise measurements of process or wall-clock time. For example, see time.perf_counter().