将值附加到DataFrame行使用lambda if else。

huangapple
go评论99阅读模式
英文:

Append value to df row using lambda if else

问题

你好，专家们，我在尝试将值附加到DataFrame的新列时遇到了问题，需要通过比较其他列的值来实现。我搜索了相关的问题，但没有找到合适的答案。我是Python的新手。

我想通过比较DataFrame的其他列的最后4个值，使用df.apply和lambda来在每个元素的最后一行附加一个新的DataFrame列。在这方面的帮助将不胜感激。

我有以下DataFrame：

  1.     Symbol    open      close      sig
  2. 0   APPL      153.60    152.90      0
  3. 1   APPL      152.90    153.55      1
  4. 2   APPL      153.55    152.00      0
  5. 3   APPL      152.00    153.50      1
  6. 4   APPL      153.50    154.10      1
  8. 5   TSLA      193.00    192.10      0
  9. 6   TSLA      192.10    191.50      0
  10. 7   TSLA      191.50    192.90      1
  11. 8   TSLA      192.90    192.45      0
  12. 9   TSLA      192.45    191.10      0

我想要比较df['sig']列，除了第1行，以此类推，对于所有的股票，如果sig列的最后4个值为APPL是1011，那么在第4行应该附加为1（即每支股票的最后一行）。如果TSLA的sig列的最后4个值为0100，那么在第9行应该附加为0。

通过使用lambda函数或df.npwhere等方法...

谢谢！

期望的结果如下：

  1.     Symbol    open      close      sig    signal
  2. 0   APPL      153.60    152.90      0      NaN
  3. 1   APPL      152.90    153.55      1      NaN
  4. 2   APPL      152.75    152.00      0      NaN
  5. 3   APPL      153.00    153.50      1      NaN
  6. 4   APPL      153.50    154.10      1      1
  8. 5   TSLA      193.00    192.10      0      NaN
  9. 6   TSLA      192.10    191.50      0      NaN
  10. 7   TSLA      191.50    192.90      1      NaN
  11. 8   TSLA      192.90    192.45      0      NaN
  12. 9   TSLA      192.45    191.10      0      0
英文:

hello experts i was struck in appending value to new column of df by comparing other column values. i searched related questions but not found suitable answer. I'm newbie to python

I want to append new column of df at last row of each element by comparing last 4 values of df other column using df.apply lambda help in this regard highly appreciated.

I have following df:

  1.     Symbol    open      close      sig
  2. 0   APPL      153.60    152.90      0
  3. 1   APPL      152.90    153.55      1
  4. 2   APPL      153.55    152.00      0
  5. 3   APPL      152.00    153.50      1
  6. 4   APPL      153.50    154.10      1
  8. 5   TSLA      193.00    192.10      0
  9. 6   TSLA      192.10    191.50      0
  10. 7   TSLA      191.50    192.90      1
  11. 8   TSLA      192.90    192.45      0
  12. 9   TSLA      192.45    191.10      0

i want to compare df['sig'] column except 1 row and so on for all stocks, if sig column last 4 values APPL is 1011 then df ['signal'] at row 4 it should be appended as 1 (i.e. last row of each stock) if last 4 values df['sig'] of TSLA is 0100 then at 9 row df['signal'] to be appended as 0

by using lambda or df.npwhere etc...

Thanks!

expected this:

  1.     Symbol    open      close      sig    signal
  2. 0   APPL      153.60    152.90      0      NaN
  3. 1   APPL      152.90    153.55      1      NaN
  4. 2   APPL      152.75    152.00      0      NaN
  5. 3   APPL      153.00    153.50      1      NaN
  6. 4   APPL      153.50    154.10      1      1
  8. 5   TSLA      193.00    192.10      0      NaN
  9. 6   TSLA      192.10    191.50      0      NaN
  10. 7   TSLA      191.50    192.90      1      NaN
  11. 8   TSLA      192.90    192.45      0      NaN
  12. 9   TSLA      192.45    191.10      0      0

答案1

得分: 1

以下是您要翻译的内容：

You can use a groupby.rolling:

  1. df['signal'] = (df.groupby('Symbol')['sig'].rolling(4)
  2.                   .apply(lambda x: x.eq([1, 0, 1, 1]).all()).droplevel(0)
  3.                )

Output:

  1.   Symbol    open   close  sig  signal
  2. 0   APPL  153.60  152.90    0     NaN
  3. 1   APPL  152.90  153.55    1     NaN
  4. 2   APPL  152.75  152.00    0     NaN
  5. 3   APPL  153.00  153.50    1     0.0
  6. 4   APPL  153.50  154.10    1     1.0
  7. 5   TSLA  193.00  192.10    0     NaN
  8. 6   TSLA  192.10  191.50    0     NaN
  9. 7   TSLA  191.50  192.90    1     NaN
  10. 8   TSLA  192.90  192.45    0     0.0
  11. 9   TSLA  192.45  191.10    0     0.0

mapping custom values for each group:

  1. mapper = {'APPL': ([1, 0, 1, 1], 1), 'TSLA': ([0, 1, 0, 0], 0)}
  3. df['signal'] = (
  4.     df.groupby('Symbol')['sig']
  5.       .apply(lambda g: g.rolling(4)
  6.                         .apply(lambda x: mapper[g.name][1]
  7.                                if x.tolist() == mapper[g.name][0]
  8.                                else float('nan')
  9.                               )
  10.             )
  11. )

Output:

  1.   Symbol    open   close  sig  signal
  2. 0   APPL  153.60  152.90    0     NaN
  3. 1   APPL  152.90  153.55    1     NaN
  4. 2   APPL  153.55  152.00    0     NaN
  5. 3   APPL  152.00  153.50    1     NaN
  6. 4   APPL  153.50  154.10    1     1.0
  7. 5   TSLA  193.00  192.10    0     NaN
  8. 6   TSLA  192.10  191.50    0     NaN
  9. 7   TSLA  191.50  192.90    1     NaN
  10. 8   TSLA  192.90  192.45    0     NaN
  11. 9   TSLA  192.45  191.10    0     0.0
英文:

You can use a groupby.rolling:

  1. df['signal'] = (df.groupby('Symbol')['sig'].rolling(4)
  2.                   .apply(lambda x: x.eq([1, 0, 1, 1]).all()).droplevel(0)
  3.                )

Output:

  1.   Symbol    open   close  sig  signal
  2. 0   APPL  153.60  152.90    0     NaN
  3. 1   APPL  152.90  153.55    1     NaN
  4. 2   APPL  152.75  152.00    0     NaN
  5. 3   APPL  153.00  153.50    1     0.0
  6. 4   APPL  153.50  154.10    1     1.0
  7. 5   TSLA  193.00  192.10    0     NaN
  8. 6   TSLA  192.10  191.50    0     NaN
  9. 7   TSLA  191.50  192.90    1     NaN
  10. 8   TSLA  192.90  192.45    0     0.0
  11. 9   TSLA  192.45  191.10    0     0.0

mapping custom values for each group:

  1. mapper = {'APPL': ([1, 0, 1, 1], 1), 'TSLA': ([0, 1, 0, 0], 0)}
  3. df['signal'] = (
  4.     df.groupby('Symbol')['sig']
  5.       .apply(lambda g: g.rolling(4)
  6.                         .apply(lambda x: mapper[g.name][1]
  7.                                if x.tolist() == mapper[g.name][0]
  8.                                else float('nan')
  9.                               )
  10.             )
  11. )

Output:

  1.   Symbol    open   close  sig  signal
  2. 0   APPL  153.60  152.90    0     NaN
  3. 1   APPL  152.90  153.55    1     NaN
  4. 2   APPL  153.55  152.00    0     NaN
  5. 3   APPL  152.00  153.50    1     NaN
  6. 4   APPL  153.50  154.10    1     1.0
  7. 5   TSLA  193.00  192.10    0     NaN
  8. 6   TSLA  192.10  191.50    0     NaN
  9. 7   TSLA  191.50  192.90    1     NaN
  10. 8   TSLA  192.90  192.45    0     NaN
  11. 9   TSLA  192.45  191.10    0     0.0

huangapple
  • 本文由 发表于 2023年2月14日 21:51:57
  • 转载请务必保留本文链接：https://go.coder-hub.com/75448817.html
  • dataframe
  • pandas
  • 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