python使用__slots__讓你的代碼更加節(jié)省內(nèi)存
前言
在默認(rèn)情況下,Python的新類和舊類的實(shí)例都有一個(gè)字典來存儲(chǔ)屬性值。這對(duì)于那些沒有實(shí)例屬性的對(duì)象來說太浪費(fèi)空間了,當(dāng)需要?jiǎng)?chuàng)建大量實(shí)例的時(shí)候,這個(gè)問題變得尤為突出。
因此這種默認(rèn)的做法可以通過在新式類中定義了一個(gè)__slots__屬性從而得到了解決。__slots__聲明中包含若干實(shí)例變量,并為每個(gè)實(shí)例預(yù)留恰好足夠的空間來保存每個(gè)變量,因此沒有為每個(gè)實(shí)例都創(chuàng)建一個(gè)字典,從而節(jié)省空間。
本文主要介紹了關(guān)于python使用__slots__讓你的代碼更加節(jié)省內(nèi)存的相關(guān)內(nèi)容,分享出來供大家參考學(xué)習(xí),下面話不多說了,來一起看看詳細(xì)的介紹吧
現(xiàn)在來說說python中dict為什么比list浪費(fèi)內(nèi)存?
和list相比,dict 查找和插入的速度極快,不會(huì)隨著key的增加而增加;dict需要占用大量的內(nèi)存,內(nèi)存浪費(fèi)多。
而list查找和插入的時(shí)間隨著元素的增加而增加;占用空間小,浪費(fèi)的內(nèi)存很少。
python解釋器是Cpython,這兩個(gè)數(shù)據(jù)結(jié)構(gòu)應(yīng)該對(duì)應(yīng)C的哈希表和數(shù)組。因?yàn)楣1硇枰~外內(nèi)存記錄映射關(guān)系,而數(shù)組只需要通過索引就能計(jì)算出下一個(gè)節(jié)點(diǎn)的位置,所以哈希表占用的內(nèi)存比數(shù)組大,也就是dict比list占用的內(nèi)存更大。
如果想更加詳細(xì)了解,可以查看C的源代碼。python官方鏈接:https://www.python.org/downloads/source/
如下代碼是我從python官方截取的代碼片段:
List 源碼:
typedef struct { PyObject_VAR_HEAD /* Vector of pointers to list elements. list[0] is ob_item[0], etc. */ PyObject **ob_item; /* ob_item contains space for 'allocated' elements. The number * currently in use is ob_size. * Invariants: * 0 <= ob_size <= allocated * len(list) == ob_size * ob_item == NULL implies ob_size == allocated == 0 * list.sort() temporarily sets allocated to -1 to detect mutations. * * Items must normally not be NULL, except during construction when * the list is not yet visible outside the function that builds it. */ Py_ssize_t allocated; } PyListObject;
Dict源碼:
/* PyDict_MINSIZE is the minimum size of a dictionary. This many slots are * allocated directly in the dict object (in the ma_smalltable member). * It must be a power of 2, and at least 4. 8 allows dicts with no more * than 5 active entries to live in ma_smalltable (and so avoid an * additional malloc); instrumentation suggested this suffices for the * majority of dicts (consisting mostly of usually-small instance dicts and * usually-small dicts created to pass keyword arguments). */ #define PyDict_MINSIZE 8 typedef struct { /* Cached hash code of me_key. Note that hash codes are C longs. * We have to use Py_ssize_t instead because dict_popitem() abuses * me_hash to hold a search finger. */ Py_ssize_t me_hash; PyObject *me_key; PyObject *me_value; } PyDictEntry; /* To ensure the lookup algorithm terminates, there must be at least one Unused slot (NULL key) in the table. The value ma_fill is the number of non-NULL keys (sum of Active and Dummy); ma_used is the number of non-NULL, non-dummy keys (== the number of non-NULL values == the number of Active items). To avoid slowing down lookups on a near-full table, we resize the table when it's two-thirds full. */ typedef struct _dictobject PyDictObject; struct _dictobject { PyObject_HEAD Py_ssize_t ma_fill; /* # Active + # Dummy */ Py_ssize_t ma_used; /* # Active */ /* The table contains ma_mask + 1 slots, and that's a power of 2. * We store the mask instead of the size because the mask is more * frequently needed. */ Py_ssize_t ma_mask; /* ma_table points to ma_smalltable for small tables, else to * additional malloc'ed memory. ma_table is never NULL! This rule * saves repeated runtime null-tests in the workhorse getitem and * setitem calls. */ PyDictEntry *ma_table; PyDictEntry *(*ma_lookup)(PyDictObject *mp, PyObject *key, long hash); PyDictEntry ma_smalltable[PyDict_MINSIZE]; };
PyObject_HEAD 源碼:
#ifdef Py_TRACE_REFS /* Define pointers to support a doubly-linked list of all live heap objects. */ #define _PyObject_HEAD_EXTRA \ struct _object *_ob_next; \ struct _object *_ob_prev; #define _PyObject_EXTRA_INIT 0, 0, #else #define _PyObject_HEAD_EXTRA #define _PyObject_EXTRA_INIT #endif /* PyObject_HEAD defines the initial segment of every PyObject. */ #define PyObject_HEAD \ _PyObject_HEAD_EXTRA \ Py_ssize_t ob_refcnt; \ struct _typeobject *ob_type;
PyObject_VAR_HEAD 源碼:
/* PyObject_VAR_HEAD defines the initial segment of all variable-size * container objects. These end with a declaration of an array with 1 * element, but enough space is malloc'ed so that the array actually * has room for ob_size elements. Note that ob_size is an element count, * not necessarily a byte count. */ #define PyObject_VAR_HEAD \ PyObject_HEAD \ Py_ssize_t ob_size; /* Number of items in variable part */
現(xiàn)在知道了dict為什么比list 占用的內(nèi)存空間更大。接下來如何讓你的類更加的節(jié)省內(nèi)存。
其實(shí)有兩種解決方案:
第一種是使用__slots__ ;另外一種是使用Collection.namedtuple 實(shí)現(xiàn)。
首先用標(biāo)準(zhǔn)的方式寫一個(gè)類:
#!/usr/bin/env python class Foobar(object): def __init__(self, x): self.x = x @profile def main(): f = [Foobar(42) for i in range(1000000)] if __name__ == "__main__": main()
然后,創(chuàng)建一個(gè)類Foobar(),然后實(shí)例化100W次。通過@profile查看內(nèi)存使用情況。
運(yùn)行結(jié)果:
該代碼共使用了372M內(nèi)存。
接下來通過__slots__代碼實(shí)現(xiàn)該代碼:
#!/usr/bin/env python class Foobar(object): __slots__ = 'x' def __init__(self, x): self.x = x @profile def main(): f = [Foobar(42) for i in range(1000000)] if __name__ == "__main__": main()
運(yùn)行結(jié)果:
使用__slots__使用了91M內(nèi)存,比使用__dict__存儲(chǔ)屬性值節(jié)省了4倍。
其實(shí)使用collection模塊的namedtuple也可以實(shí)現(xiàn)__slots__相同的功能。namedtuple其實(shí)就是繼承自tuple,同時(shí)也因?yàn)開_slots__的值被設(shè)置成了一個(gè)空tuple以避免創(chuàng)建__dict__。
看看collection是如何實(shí)現(xiàn)的:
collection 和普通創(chuàng)建類方式相比,也節(jié)省了不少的內(nèi)存。所在在確定類的屬性值固定的情況下,可以使用__slots__方式對(duì)內(nèi)存進(jìn)行優(yōu)化。但是這項(xiàng)技術(shù)不應(yīng)該被濫用于靜態(tài)類或者其他類似場(chǎng)合,那不是python程序的精神所在。
總結(jié)
以上就是這篇文章的全部?jī)?nèi)容了,希望本文的內(nèi)容對(duì)大家的學(xué)習(xí)或者工作具有一定的參考學(xué)習(xí)價(jià)值,如果有疑問大家可以留言交流,謝謝大家對(duì)腳本之家的支持。
相關(guān)文章
pycharm設(shè)置當(dāng)前工作目錄的操作(working directory)
今天小編就為大家分享一篇pycharm設(shè)置當(dāng)前工作目錄的操作(working directory),具有很好的參考價(jià)值,希望對(duì)大家有所幫助。一起跟隨小編過來看看吧2020-02-02使用50行Python代碼從零開始實(shí)現(xiàn)一個(gè)AI平衡小游戲
本文會(huì)為大家展示機(jī)器學(xué)習(xí)專家 Mike Shi 如何用 50 行 Python 代碼創(chuàng)建一個(gè) AI,使用增強(qiáng)學(xué)習(xí)技術(shù),玩耍一個(gè)保持桿子平衡的小游戲。本文給大家?guī)韺?shí)現(xiàn)思路及簡(jiǎn)單代碼,感興趣的朋友跟隨小編一起看看吧2018-11-11如何基于Python實(shí)現(xiàn)一個(gè)慶祝國(guó)慶節(jié)的小程序
這篇文章主要介紹了如何基于Python實(shí)現(xiàn)一個(gè)慶祝國(guó)慶節(jié)的小程序,增加了互動(dòng)選擇祝福語、查詢信息、播放背景音樂及趣味小測(cè)驗(yàn)等功能,使用tkinter增強(qiáng)GUI,提升用戶互動(dòng)體驗(yàn),需要的朋友可以參考下2024-09-09Python實(shí)現(xiàn)端口復(fù)用實(shí)例代碼
這篇文章主要介紹了Python實(shí)現(xiàn)端口復(fù)用實(shí)例代碼,需要的朋友可以參考下2014-07-07