从 CSV 中，将相似的数据组合在一起？

Question

我有一些数据要组合在一起：

Serial_Num     Latitude Longitude
1950004S11059   -11.1   59.1
1950004S11059   -11.6   57.8
1950004S11059   -12.4   56
1950004S11059   -13.2   54.6
1950004S11059   -13.8   53.8
1950004S11059   -14.8   52.7
1950004S11059   -15.9   52
1950004S11059   -18.3   52.4
1950004S11059   -20     54
1950004S11059   -22.1   55.9
1950004S11059   -26.2   59.8
1950012S14150   -14     146.9
1950012S14150   -14.4   145.8
1950012S14150   -14.9   145.4
1950012S14150   -15.8   145.6
1950012S14150   -18.9   149.1
1950012S14150   -22.3   152.5
1950013S14139   -16     139
1950013S14139   -16.3   139

很简单，对于每个唯一的Serial_Num，我想要坐标。我期待类似的东西：

1950004S11059: {"GPS": (-11.1 , 59.1) , (-11.6, 57.8) , (-12.4, 56), ..., (-26.2, 59.8)}

然后我可以循环遍历每个Serial_Num 的GPS 坐标并进行绘图。

我有一些我在其他地方使用过的脚本，但主要依赖于用于设置字典的 .csv 数据，以 Serial_Num 为键。

但是csv中的数据是顺序的，顺序很重要。

对于每个Serial_Num，有什么方法可以按照它们在 CSV 中的顺序输出坐标列表？

编辑：我现在正在研究 Pandas，因为它有一个 groupBy 方法可能会有所帮助。

Answer 1

这是一种方式。下面对每个步骤进行细分。

import pandas as pd

df = pd.read_csv('file.csv', delim_whitespace=True)
df['GPS'] = list(zip(df.Latitude, df.Longitude))
df.groupby('Serial_Num')['GPS'].apply(list).to_dict()

读取数据

df = pd.read_csv('file.csv', delim_whitespace=True)

#        Serial_Num  Latitude  Longitude
# 0   1950004S11059     -11.1       59.1
# 1   1950004S11059     -11.6       57.8
# 2   1950004S11059     -12.4       56.0
# 3   1950004S11059     -13.2       54.6
# 4   1950004S11059     -13.8       53.8
# 5   1950004S11059     -14.8       52.7

制作元组列

df['GPS'] = list(zip(df.Latitude, df.Longitude))

#        Serial_Num  Latitude  Longitude             GPS
# 0   1950004S11059     -11.1       59.1   (-11.1, 59.1)
# 1   1950004S11059     -11.6       57.8   (-11.6, 57.8)
# 2   1950004S11059     -12.4       56.0   (-12.4, 56.0)
# 3   1950004S11059     -13.2       54.6   (-13.2, 54.6)
# 4   1950004S11059     -13.8       53.8   (-13.8, 53.8)
# 5   1950004S11059     -14.8       52.7   (-14.8, 52.7)

创建字典

df.groupby('Serial_Num')['GPS'].apply(list).to_dict()

# {'1950004S11059': [(-11.1, 59.100000000000001),
#                    (-11.6, 57.799999999999997),
#                    (-12.4, 56.0),
#                    (-13.199999999999999, 54.600000000000001),
#                    (-13.800000000000001, 53.799999999999997),
#                    (-14.800000000000001, 52.700000000000003),
#                    (-15.9, 52.0),
#                    (-18.300000000000001, 52.399999999999999),
#                    (-20.0, 54.0),
#                    (-22.100000000000001, 55.899999999999999),
#                    (-26.199999999999999, 59.799999999999997)],
#  '1950012S14150': [(-14.0, 146.90000000000001),
#                    (-14.4, 145.80000000000001),
#                    (-14.9, 145.40000000000001),
#                    (-15.800000000000001, 145.59999999999999),
#                    (-18.899999999999999, 149.09999999999999),
#                    (-22.300000000000001, 152.5)],
#  '1950013S14139': [(-16.0, 139.0), (-16.300000000000001, 139.0)]}

Answer 2

给定

一个名为foo.csv的文件：

Serial_Num     Latitude Longitude
1950004S11059   -11.1   59.1
1950004S11059   -11.6   57.8
1950004S11059   -12.4   56
1950004S11059   -13.2   54.6
1950004S11059   -13.8   53.8
1950004S11059   -14.8   52.7
1950004S11059   -15.9   52
1950004S11059   -18.3   52.4
1950004S11059   -20     54
1950004S11059   -22.1   55.9
1950004S11059   -26.2   59.8
1950012S14150   -14     146.9
1950012S14150   -14.4   145.8
1950012S14150   -14.9   145.4
1950012S14150   -15.8   145.6
1950012S14150   -18.9   149.1
1950012S14150   -22.3   152.5
1950013S14139   -16     139
1950013S14139   -16.3   139

还有一些代码将数据解析为（序列，坐标）的元组：

import csv
import collections as ct


def read_file(fname):
    with open(fname) as f:
        reader = csv.reader(f)
        next(reader)
        for line in reader:
            #line = [x for x in line[0].split(" ") if x]
            yield line[0], tuple(map(float, line[1:]))

代码

我们构建了defaultdicts的嵌套：

data = ct.defaultdict(dict)
for serial, coords in (read_file("foo.csv")):
    if serial not in data:
        dd = ct.defaultdict(list)
    dd["GPS"].append(coords)
    data[serial] = dd

dict(data)

输出

{'1950004S11059': defaultdict(list,
             {'GPS': [
               (-11.1, 59.1),
               (-11.6, 57.8),
               (-12.4, 56.0),
               (-13.2, 54.6),
               (-13.8, 53.8),
               (-14.8, 52.7),
               (-15.9, 52.0),
               (-18.3, 52.4),
               (-20.0, 54.0),
               (-22.1, 55.9),
               (-26.2, 59.8)]}),
 '1950012S14150': defaultdict(list,
             {'GPS': [
               (-14.0, 146.9),
               (-14.4, 145.8),
               (-14.9, 145.4),
               (-15.8, 145.6),
               (-18.9, 149.1),
               (-22.3, 152.5)]}),
 '1950013S14139': defaultdict(list, 
             {'GPS': [
               (-16.0, 139.0), 
               (-16.3, 139.0)]})}