004.python科学计算库pandas(中)

灰太狼 2021-10-06 15:16 428阅读 0赞

### 测试数据 [titanic\_train.csv][titanic_train.csv] ###

--------------------

### isnull ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # Pandas库使用NaN(非数字)表示缺失值 # 我们可以使用pandas.isnull()函数，它获取一个pandas series并返回一系列的True和False age = titanic_survival["Age"] # 使用loc获取的数据时的切片，包括两端的索引对应的数据 print(age.loc[0:6]) print("--------------------------") age_is_null = pandas.isnull(age) # 对普通列表数据的切片，不包括右端的索引对应的数据 print(age_is_null[0:6]) print("--------------------------") print(type(age_is_null)) print("--------------------------") print(len(age), len(age_is_null))

![pycharm64_2018-09-14_14-43-02.png][]

--------------------

### nan ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # 其结果是mean_age将是nan。这是因为我们对空值所做的任何计算都会得到空值 age = titanic_survival["Age"] print(sum(age)) print("--------------------------") mean_age = sum(age) / len(age) print(mean_age) print("--------------------------") # 在计算平均值之前，我们必须过滤掉遗漏的值 age_is_null = pandas.isnull(age) good_ages = age[age_is_null == False] correct_mean_age = sum(good_ages) / len(good_ages) print(correct_mean_age) print("--------------------------") # 丢失的数据是如此常见，以至于许多pandas方法会自动为其过滤 correct_mean_age = age.mean() print(correct_mean_age)

![pycharm64_2018-09-14_14-55-07.png][]

--------------------

### mean ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # 每节课平均票价 passenger_classes = [1, 2, 3] fares_by_class = { } for this_class in passenger_classes: # 获取所有满足titanic_survival["Pclass"] == this_class 的数据 pclass_rows = titanic_survival[titanic_survival["Pclass"] == this_class] # 获取符合this_class的数据的Fare列 pclass_fares = pclass_rows["Fare"] # 求平均数并赋值到字典中 fares_by_class[this_class] = pclass_fares.mean() print(fares_by_class)

![pycharm64_2018-09-14_15-05-59.png][]

--------------------

### pivot\_table ###

import pandas import numpy titanic_survival = pandas.read_csv("titanic_train.csv") # pivot_table 创建一个电子表格样式的数据透视表。pivot表中的级别将存储在结果DataFrame的索引和列上的多索引对象(层次索引)中 # index 告诉方法按哪个列分组 # values 是我们要应用计算的列(可选地聚合列) # aggfunc 指定我们要执行的计算 default numpy.mean 沿着指定的轴计算算术平均数 passenger_survival = titanic_survival.pivot_table(index="Pclass", values="Survived", aggfunc=numpy.mean) print(passenger_survival) print("------------------------") passenger_age = titanic_survival.pivot_table(index="Pclass", values="Age") print(passenger_age) print("------------------------") port_stats = titanic_survival.pivot_table(index="Embarked", values=["Fare", "Survived"], aggfunc=numpy.sum) print(port_stats)

![pycharm64_2018-09-14_15-15-56.png][]

import pandas def generate_age_label(row): age = row["Age"] if pandas.isnull(age): return "unknown" elif age < 18: return "minor" else: return "adult" titanic_survival = pandas.read_csv("titanic_train.csv") # 从每列返回第100项 # apply 沿着DataFrame的轴应用一个函数。 # axis : {0 or 'index', 1 or 'columns'}, default 0 # 0 or 'index': apply function to each column. 每列 # 1 or 'columns': apply function to each row. 每行 age_labels = titanic_survival.apply(generate_age_label, axis=1) titanic_survival['age_labels666'] = age_labels age_group_survival = titanic_survival.pivot_table(index="age_labels666", values="Survived") print(age_group_survival)

![pycharm64_2018-09-14_17-32-34.png][]

--------------------

### dropna ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # dropna 删除缺失值 # axis = 0或'index': 删除包含缺失值的行 # axis = 1或'columns': 删除包含缺失值的列 # subset 像数组一样，可选的标签沿着要考虑的其他轴，例如，如果要删除行，这些将是要包含的列的列表。 # how : {'any', 'all'} # 'any' : 如果存在任何NA值，删除该行或列。 # 'all' : 如果所有值都是NA，则删除该行或列。 drop_na_columns = titanic_survival.dropna(axis=1) print(drop_na_columns[0:3]) print("-------------------------------------------") new_titanic_survival = titanic_survival.dropna(axis=0, subset=["Age", "Sex"]) print(len(new_titanic_survival)) print("-------------------------------------------") print(new_titanic_survival[0:3])

![pycharm64_2018-09-14_15-30-30.png][]

--------------------

### loc ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # 获取第84行数据的Age列的值 (loc索引下标从0开始) row_index_83_age = titanic_survival.loc[83, "Age"] # 获取第767行数据的Pclass列的值 (loc索引下标从0开始) row_index_766_pclass = titanic_survival.loc[766, "Pclass"] print(row_index_83_age) print(row_index_766_pclass)

![pycharm64_2018-09-14_15-46-00.png][]

--------------------

### reset\_index ###

import pandas titanic_survival = pandas.read_csv("titanic_train.csv") # 按照Age列进行升序排序 new_titanic_survival = titanic_survival.sort_values("Age", ascending=False) print(new_titanic_survival[0:3]) print("-------------------------------") # reset_index 对于具有多层索引的DataFrame，返回新的DataFrame，并在索引名下的列中标记信息， # 如果没有，默认为'level_0'、'level_1'等。 # 对于标准索引，将使用索引名称(如果设置)， # 否则将使用默认的“index”或“level_0”(如果已经使用了“index”)。 # drop : boolean, default False 不要尝试在dataframe列中插入索引。这会将索引重置为默认整数索引。 titanic_reindexed = new_titanic_survival.reset_index(drop=True) titanic_reindexed2 = new_titanic_survival.reset_index() print(titanic_reindexed2.iloc[0:3]) print("-------------------------------") print(titanic_reindexed.iloc[0:3])

![pycharm64_2018-09-14_16-17-52.png][]

--------------------

### apply ###

import pandas def hundredth_row(column): # 提取第100项 hundredth_item = column.iloc[99] return hundredth_item titanic_survival = pandas.read_csv("titanic_train.csv") # 从每列返回第100项 # apply 沿着DataFrame的轴应用一个函数。 # axis : {0 or 'index', 1 or 'columns'}, default 0 # 0 or 'index': apply function to each column. 每列 # 1 or 'columns': apply function to each row. 每行 hundredth_data = titanic_survival.apply(hundredth_row) print(hundredth_data)

![pycharm64_2018-09-14_17-13-36.png][]

import pandas def not_null_count(column): column_null = pandas.isnull(column) null = column[column_null] return len(null) titanic_survival = pandas.read_csv("titanic_train.csv") # 从每列返回第100项 # apply 沿着DataFrame的轴应用一个函数。 # axis : {0 or 'index', 1 or 'columns'}, default 0 # 0 or 'index': apply function to each column. 每列 # 1 or 'columns': apply function to each row. 每行 column_null_count = titanic_survival.apply(not_null_count) print(column_null_count)

![pycharm64_2018-09-14_17-15-03.png][]

import pandas def which_class(row): pclass = row['Pclass'] if pandas.isnull(pclass): return "Unknown" elif pclass == 1: return "First Class" elif pclass == 2: return "Second Class" elif pclass == 3: return "Third Class" titanic_survival = pandas.read_csv("titanic_train.csv") # 从每列返回第100项 # apply 沿着DataFrame的轴应用一个函数。 # axis : {0 or 'index', 1 or 'columns'}, default 0 # 0 or 'index': apply function to each column. 每列 # 1 or 'columns': apply function to each row. 每行 classes = titanic_survival.apply(which_class, axis=1) print(classes)

![mspaint_2018-09-14_17-17-10.png][]

import pandas def is_minor(row): if row["Age"] < 18: return True else: return False titanic_survival = pandas.read_csv("titanic_train.csv") # 从每列返回第100项 # apply 沿着DataFrame的轴应用一个函数。 # axis : {0 or 'index', 1 or 'columns'}, default 0 # 0 or 'index': apply function to each column. 每列 # 1 or 'columns': apply function to each row. 每行 minors = titanic_survival.apply(is_minor, axis=1) print(minors)

![mspaint_2018-09-14_17-20-06.png][]

![mspaint_2018-09-14_17-23-42.png][]

--------------------

[titanic_train.csv]: http://upyun.qubianzhong.cn/file/python_pandas/titanic_train.csv
[pycharm64_2018-09-14_14-43-02.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_14-43-02.png
[pycharm64_2018-09-14_14-55-07.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_14-55-07.png
[pycharm64_2018-09-14_15-05-59.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_15-05-59.png
[pycharm64_2018-09-14_15-15-56.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_15-15-56.png
[pycharm64_2018-09-14_17-32-34.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_17-32-34.png
[pycharm64_2018-09-14_15-30-30.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_15-30-30.png
[pycharm64_2018-09-14_15-46-00.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_15-46-00.png
[pycharm64_2018-09-14_16-17-52.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_16-17-52.png
[pycharm64_2018-09-14_17-13-36.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_17-13-36.png
[pycharm64_2018-09-14_17-15-03.png]: http://upyun.qubianzhong.cn/picture/pycharm64_2018-09-14_17-15-03.png
[mspaint_2018-09-14_17-17-10.png]: http://upyun.qubianzhong.cn/picture/mspaint_2018-09-14_17-17-10.png
[mspaint_2018-09-14_17-20-06.png]: http://upyun.qubianzhong.cn/picture/mspaint_2018-09-14_17-20-06.png
[mspaint_2018-09-14_17-23-42.png]: http://upyun.qubianzhong.cn/picture/mspaint_2018-09-14_17-23-42.png