How to compare columns in pandas

Подписаться 11 тыс.

Просмотров 18 тыс.

50% 1

In this video we go over how to compare columns in a pandas DataFrame. We also go over some issues to look out for when compare column values.
We create a new column with Boolean values to indicate everywhere that our comparison was satisfied.
We also discuss how to use arithmetic with Boolean values.
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
$15 off Annual Dataquest subscription
app.dataquest.io/referral-signup/qybqz3r8/
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
Did you find this video helpful? Consider subscribing for weekly tips, tricks, and tutorials.
/ @chartexplorers
0:00 Intro
0:08 Setup
0:27 Compare Columns
0:53 Operators
1:04 Gottcha
127 Arithmetic & Boolean

Наука

Опубликовано:

13 июл 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 24

@mrmuranga 3 года назад

awesome.....keep making these informative short videos

@ChartExplorers 3 года назад

Will do! Thanks for your encouragement.

@studyhub3950 3 года назад

very good helping and informative and short. keep it up.

@ChartExplorers 3 года назад

Thank you!

@shobhitasohal2853 2 года назад

hey. Can we use if statement to compare 2 columns in a new column & return certain value other than just true or false?

@jayineedesai7322 2 года назад

Hi @chart Explorers , i have one doubt.. when i used the same method shown here ..it is giving TypeError: unhashable type: 'Series' can you please tell me how to resolve this?

@tanliz765 3 года назад

Thank you!

@ChartExplorers 3 года назад

You're Welcome! :)

@vinayakchikkorde8151 2 года назад

I have the source file and target file. so in that, I have to compare 140 columns and show the result if it matches or not. for example, there is a column as Country1 in source and in target as Country2. to compare that i will use if(source['country1]==target['country2])return True else return false. to compare 140+ columns it will take time to compare 140 columns. so how can I solve this?

@sanahnahk7312 3 года назад

thanks !

@ChartExplorers 3 года назад

You're Welcome!

@salman_shaikh 2 года назад

unable to compare same thing with float values. getting incorrect result

@bitwamet 2 года назад

How to add multiple condition

@gussta1 3 года назад

Instead of it saying True / False, is there a way to make say any type of "string value", like high / low or full / empty...anything other than True / False

@ChartExplorers 3 года назад

This is a little cluncky df['boo'] = (df['col1'] < df['col2']).apply(lambda x: 'Less than' if True else 'Greater than eqaul to') or you could use something slightly more performant condlist = [df['col1'] < df['col2'],df['col1'] >= df['col2']] choicelist = ['Less than', 'Greater than'] df['boo'] = np.select(condlist, choicelist) Let me know if you want an explanation (I won't be able to give it for the next 10 hours or so)

@khushbugupta6970 2 года назад

how to compare one column with the other 16 columns? These 16 columns are not next to each other.

@miss_tech 2 года назад

ValueError: Can only compare identically-labeled Series objects

@hellofrombd1960 3 года назад

How 0 5 5 5 5 5 Comes???? Don't understand. If False means 0 then df['boo']+1 would be 0+1=1 Then 1+1=2 2 2 2

@ChartExplorers 3 года назад

Good catch. This was an editing mistake, I had provided more examples in my original explanation, but edited them out. If you notice at 1:46 seconds we get the expected answer (2) and then I cut out my other examples in editing. I'll see what I can do to fix this. Thanks.

@noob_coder001 2 года назад

Comapare two colums get the records from one column..EXAMPLE A column values == Bcolummn values same print(Common values )😭😭

@BhanuLekha 3 года назад

How to code for how many trues r there

@ChartExplorers 3 года назад

Do you want to count the number of True values in a column? df['col_name'].sum()

@BhanuLekha 3 года назад

@@ChartExplorers s I need no.of true values in the column after comparing .thank u

@ChartExplorers 3 года назад

@@BhanuLekha A neat feature of boolean values (True and False) is that they are considered as 1's and 0's. So you can use mathematical operations on them just like a floats or ints. If all you need is the sum you can : (df['Col_1'] > df['Col_2']).sum() Or you could create a new column (like in the video) and then sum the column. df['new_col'] = df['Col_1'] > df['Col_2'] df['new_col'].sum() This will return all the number of true values. If you wanted the number of False values len(df) - (df['Col_1'] > df['Col_2']).sum()