计算定义错误？

huangapple
go评论108阅读模式
英文:

Definition calculating incorrectly?

问题

下面是代码部分的翻译结果，只返回翻译好的内容，不包括其他内容：

  1. # 带有分数的数据帧
  2. # 应返回该分数的浮点值。
  4. df:
  5.    Unnamed: 0  country      league home_odds draw_odds away_odds             datetime      home_team           away_team  home_score  away_score
  6. 0      381084  Iceland  League Cup    167/50   329/100    63/100  2016-02-16 19:15:00      Kopavogur  Vikingur Reykjavik           0           1
  7. 1      381085  Iceland  League Cup   463/100    173/50    47/100  2016-02-14 21:15:00           Fram            Stjarnan           0           3
  8. 2      381086  Iceland  League Cup      9/25     99/25    303/50  2016-02-14 19:15:00   KR Reykjavik              Haukar           1           1
  9. 3      381087  Iceland  League Cup      9/25   393/100   611/100  2016-02-14 19:00:00  Thor Akureyri          Leiknir F.           5           0
  10. 4      381088  Iceland  League Cup     11/25   353/100    251/50  2016-02-14 17:00:00        Akranes           Grindavik           5           0
  12. # 使用以下代码进行转换：
  13. def convert(s):
  14.     if '/' in str(s):  # 是一个分数
  15.         num, den = s.split('/')
  16.         return 1 + int(num) / int(den)
  17.     else:
  18.         return float(s)
  20. odds_cols = ['home_odds', 'draw_odds', 'away_odds']
  22. df[odds_cols] = df[odds_cols].applymap(convert)

希望这能帮助你正确返回所需的值。

关于您的最后一个问题，Pandas 将 "167/50" 识别为 float 的原因是因为它在内部执行了自动类型转换。当您使用 applymap 函数时，Pandas 会尝试将每个单元格的值解析为适当的数据类型。在这种情况下，它能够识别 "167/50" 作为分数并将其转换为浮点数。这是 Pandas 数据框的一个强大功能，可以自动识别和处理各种数据类型。

英文:

I have a dataframe with fractions.

It should return the float value of that fraction.

  1. df:
  2.    Unnamed: 0  country      league home_odds draw_odds away_odds             datetime      home_team           away_team  home_score  away_score
  3. 0      381084  Iceland  League Cup    167/50   329/100    63/100  2016-02-16 19:15:00      Kopavogur  Vikingur Reykjavik           0           1
  4. 1      381085  Iceland  League Cup   463/100    173/50    47/100  2016-02-14 21:15:00           Fram            Stjarnan           0           3
  5. 2      381086  Iceland  League Cup      9/25     99/25    303/50  2016-02-14 19:15:00   KR Reykjavik              Haukar           1           1
  6. 3      381087  Iceland  League Cup      9/25   393/100   611/100  2016-02-14 19:00:00  Thor Akureyri          Leiknir F.           5           0
  7. 4      381088  Iceland  League Cup     11/25   353/100    251/50  2016-02-14 17:00:00        Akranes           Grindavik           5           0

I am using:

  1. def convert(s):
  2.     if '/' in str(s):  # is a fraction
  3.         num, den = s.split('/')
  4.         return 1 + int(num) / int(den)
  5.     else:
  6.         return float(s)
  8. odds_cols = ['home_odds', 'draw_odds', 'away_odds']
  10. df[odds_cols] = df[odds_cols].applymap(convert)

White it returns

  1.    Unnamed: 0  country      league  home_odds  draw_odds  away_odds             datetime      home_team           away_team  home_score  away_score
  2. 0      381084  Iceland  League Cup       3.34       3.29       0.63  2016-02-16 19:15:00      Kopavogur  Vikingur Reykjavik           0           1
  3. 1      381085  Iceland  League Cup       4.63       3.46       0.47  2016-02-14 21:15:00           Fram            Stjarnan           0           3
  4. 2      381086  Iceland  League Cup       0.36       3.96       6.06  2016-02-14 19:15:00   KR Reykjavik              Haukar           1           1
  5. 3      381087  Iceland  League Cup       0.36       3.93       6.11  2016-02-14 19:00:00  Thor Akureyri          Leiknir F.           5           0
  6. 4      381088  Iceland  League Cup       0.44       3.53       5.02  2016-02-14 17:00:00        Akranes           Grindavik           5           0

It should return

  1.    Unnamed: 0  country      league  home_odds  draw_odds  away_odds             datetime      home_team           away_team  home_score  away_score
  2. 0      381084  Iceland  League Cup       4.34       4.29       1.63  2016-02-16 19:15:00      Kopavogur  Vikingur Reykjavik           0           1
  3. 1      381085  Iceland  League Cup       5.63       4.46       1.47  2016-02-14 21:15:00           Fram            Stjarnan           0           3
  4. 2      381086  Iceland  League Cup       1.36       4.96       7.06  2016-02-14 19:15:00   KR Reykjavik              Haukar           1           1
  5. 3      381087  Iceland  League Cup       1.36       4.93       7.11  2016-02-14 19:00:00  Thor Akureyri          Leiknir F.           5           0
  6. 4      381088  Iceland  League Cup       1.44       4.53       6.02  2016-02-14 17:00:00        Akranes           Grindavik           5           0

Since odds_cols have float and int values too, I cannot just apply return 1 + float(s)

What is the correct way to return the values as desired?

Also, when I type

  1. print(df.dtypes)
  4. Unnamed: 0      int64
  5. country        object
  6. league         object
  7. home_odds     float64
  8. draw_odds     float64
  9. away_odds     float64
  10. datetime       object
  11. home_team      object
  12. away_team      object
  13. home_score      int64
  14. away_score      int64
  15. dtype: object

How come pandas identifies the value 167/50 as float?

答案1

得分: 2

如果home_odds、draw_odds和away_odds中的数据以字符串格式显示，那么我可以从您的确切函数中获得您想要的结果。

我会尝试将odds_cols强制转换为字符串，例如：

  1. df['home_odds'] = df['home_odds'].astype(str)

然后查看输出。

英文:

If the data in home_odds, draw_odds and away_odds is as shown in the strings format then I get exactly what you want from your exact function.

I'd test casting the odds_cols to strings e.g.

  1. df['home_odds'] = df['home_odds'].astype(str)

and see the output.

答案2

得分: 1

你的函数可能不会在捕获分数的条件下执行：

  1. def convert(s):
  2.     if '/' in str(s):  # 是一个分数
  3.         # 试着在这里加入一个打印语句
  4.         num, den = s.split('/')
  5.         return 1 + int(num) / int(den)
  6.     else:
  7.         return float(s)

很可能你总是会进入 float(s) 的执行路径，因为框架中保留的是数字而不是字符串。

英文:

your function likely doesn't get executed on the condition to catch the fraction

  1. def convert(s):
  2.     if '/' in str(s):  # is a fraction
  3.         # TRY ADDING a print statement here 
  4.         num, den = s.split('/')
  5.         return 1 + int(num) / int(den)
  6.     else:
  7.         return float(s)

Likely you always get the float(s) execution path, all because the frame keeps the numbers, not strings.

huangapple
  • 本文由 发表于 2023年2月27日 12:23:19
  • 转载请务必保留本文链接：https://go.coder-hub.com/75576759.html
  • dataframe
  • function
  • python
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定