What is the Pandas equivalent of top_n() in dplyr? - python

What is the Pandas equivalent of top_n() in dplyr?
In R dplyr 0.8.5:
> df <- data.frame(x = c(10, 4, 1, 6, 3, 1, 6))
> df %>% top_n(2, wt=x)
x
1 10
2 6
3 6
As the dplyr documentation highlights, note that we get more than 2 values here because there's a tie: top_n() either takes all rows with a value, or none.
My attempt in Pandas 1.0.1:
df = pd.DataFrame({'x': [10, 4, 1, 6, 3, 1, 6]})
df = df.sort_values('x', ascending=False)
df.groupby('x').head(2)
Result:
x
0 10
3 6
6 6
1 4
4 3
2 1
5 1
Expected results:
x
1 10
2 6
3 6

Use parameter keep='all' in DataFrame.nlargest, sorting here is not necessary:
df = df.nlargest(2, 'x', keep='all')
print(df)
x
0 10
3 6
6 6

IUC, try series.nlargest with series.isin:
df[df['x'].isin(df['x'].nlargest(2))]#.reset_index(drop=True)
x
0 10
3 6
6 6

top_n in dplyr is superseded by slice_max/slice_min. See:
https://dplyr.tidyverse.org/reference/top_n.html
With datar in python, you can do it in a similar way:
>>> import pandas as pd
>>> from datar.all import f, slice_max
>>>
>>> df = pd.DataFrame({'x': [10, 4, 1, 6, 3, 1, 6]})
>>> df
x
<int64>
0 10
1 4
2 1
3 6
4 3
5 1
6 6
>>> df >> slice_max(n=3, order_by=f.x)
x
<int64>
0 10
3 6
6 6
Disclaimer: I am the author of the datar package.

Related

Replace specific values in a data frame with column mean

I have a dataframe and I want to replace the value 7 with the round number of mean of its columns with out other 7 in that columns. Here is a simple example:
import pandas as pd
df = pd.DataFrame()
df['a'] = [1, 2, 3]
df['b'] =[3, 0, -1]
df['c'] = [4, 7, 6]
df['d'] = [7, 7, 6]
a b c d
0 1 3 4 7
1 2 0 7 7
2 3 -1 6 6
And here is the output I want:
a b c d
0 1 3 4 2
1 2 0 3 2
2 3 -1 6 6
For example, in row 1, the mean of column c is equal to 3.33 and then its round is 3, and in column column d is equal to 2 (since we do not consider the other 7 in that column).
Can you please help me with that?
here is one way to do it
# replace 7 with np.nan
df.replace(7,np.nan, inplace=True)
# fill NaN values with the mean of the column
(df.fillna(df.apply(lambda x: x.replace(np.nan, 0)
.mean(skipna=False) ))
.round(0)
.astype(int))
a b c d
0 1 3 4 2
1 2 0 3 2
2 3 -1 6 6
temp = df.replace(to_replace=7, value=0, inplace=False).copy()
df.replace(to_replace=7, value=temp.mean().astype(int), inplace=True)

How can I extract a column from dataframe and attach it to rows while keeping other columns intact

How can I extract a column from pandas dataframe attach it to rows while keeping the other columns same.
This is my example dataset.
import pandas as pd
import numpy as np
df = pd.DataFrame({'ID': np.arange(0,5),
'sample_1' : [5,6,7,8,9],
'sample_2' : [10,11,12,13,14],
'group_id' : ["A","B","C","D","E"]})
The output I'm looking for is:
df2 = pd.DataFrame({'ID': [0, 1, 2, 3, 4, 0, 1, 2, 3, 4],
'sample_1' : [5,6,7,8,9,10,11,12,13,14],
'group_id' : ["A","B","C","D","E","A","B","C","D","E"]})
I have tried to slice the dataframe and concat using pd.concat but it was giving NaN values.
My original dataset is large.
You could do this using stack: Set the index to the columns you don't want to modify, call stack, sort by the "sample" column, then reset your index:
df.set_index(['ID','group_id']).stack().sort_values(0).reset_index([0,1]).reset_index(drop=True)
ID group_id 0
0 0 A 5
1 1 B 6
2 2 C 7
3 3 D 8
4 4 E 9
5 0 A 10
6 1 B 11
7 2 C 12
8 3 D 13
9 4 E 14
Using pd.wide_to_long:
res = pd.wide_to_long(df, stubnames='sample_', i='ID', j='group_id')
res.index = res.index.droplevel(1)
res = res.rename(columns={'sample_': 'sample_1'}).reset_index()
print(res)
ID group_id sample_1
0 0 A 5
1 1 B 6
2 2 C 7
3 3 D 8
4 4 E 9
5 0 A 10
6 1 B 11
7 2 C 12
8 3 D 13
9 4 E 14
The function you are looking for is called melt
For example:
df2 = pd.melt(df, id_vars=['ID', 'group_id'], value_vars=['sample_1', 'sample_2'], value_name='sample_1')
df2 = df2.drop('variable', axis=1)

Vectorized calculation of a column's value based on a previous value of the same column?

I have a pandas dataframe with two columns A,B as below.
I want a vectorized solution for creating a new column C where C[i] = C[i-1] - A[i] + B[i].
df = pd.DataFrame(data={'A': [10, 2, 3, 4, 5, 6], 'B': [0, 1, 2, 3, 4, 5]})
>>> df
A B
0 10 0
1 2 1
2 3 2
3 4 3
4 5 4
5 6 5
Here is the solution using for-loops:
df['C'] = df['A']
for i in range(1, len(df)):
df['C'][i] = df['C'][i-1] - df['A'][i] + df['B'][i]
>>> df
A B C
0 10 0 10
1 2 1 9
2 3 2 8
3 4 3 7
4 5 4 6
5 6 5 5
... which does the job.
But since loops are slow in comparison to vectorized calculations, I want a vectorized solution for this in pandas:
I tried to use the shift() method like this:
df['C'] = df['C'].shift(1).fillna(df['A']) - df['A'] + df['B']
but it didn't help since the shifted C column isn't updated with the calculation. It keeps its original values:
>>> df['C'].shift(1).fillna(df['A'])
0 10
1 10
2 2
3 3
4 4
5 5
and that produces a wrong result.
This can be vectorized since:
delta[i] = C[i] - C[i-1] = -A[i] +B[i]. You can get delta from A and B first, then...
calculate cumulative sum of delta (plus C[0]) to get full C
Code as follows:
delta = df['B'] - df['A']
delta[0] = 0
df['C'] = df.loc[0, 'A'] + delta.cumsum()
​
print df
A B C
0 10 0 10
1 2 1 9
2 3 2 8
3 4 3 7
4 5 4 6
5 6 5 5

Pandas index column by boolean

I want to keep columns that have 'n' or more values.
For example:
> df = pd.DataFrame({'a': [1,2,3], 'b': [1,None,4]})
a b
0 1 1
1 2 NaN
2 3 4
3 rows × 2 columns
> df[df.count()==3]
IndexingError: Unalignable boolean Series key provided
> df[:,df.count()==3]
TypeError: unhashable type: 'slice'
> df[[k for (k,v) in (df.count()==3).items() if v]]
a
0 1
1 2
2 3
Is that the best way to do this? It seems ridiculous.
You can use conditional list comprehension to generate the columns that exceed your threshold (e.g. 3). Then just select those columns from the data frame:
# Create sample DataFrame
df = pd.DataFrame({'a': [1, 2, 3, 4, 5],
'b': [1, None, 4, None, 2],
'c': [5, 4, 3, 2, None]})
>>> df_new = df[[col for col in df if df[col].count() > 3]]
Out[82]:
a c
0 1 5
1 2 4
2 3 3
3 4 2
4 5 NaN
Use count to produce a boolean index and use this as a mask for the columns:
In [10]:
df[df.columns[df.count() > 2]]
Out[10]:
a
0 1
1 2
2 3
if you want to keep columns that have 'n' or more values. for my example i am considering n value as 4
df = pd.DataFrame({'a': [1,2,3,4,6], 'b': [1,None,4,5,7],'c': [1,2,3,5,8]})
print df
a b c
0 1 1 1
1 2 NaN 2
2 3 4 3
3 4 5 5
4 6 7 8
print df[[i for i in xrange(0,len(df.columns)) if len(df.iloc[:,i]) - df.isnull().sum()[i] >4]]
a c
0 1 1
1 2 2
2 3 3
3 4 5
4 6 8

How to sort pandas dataframe from list category?

so I have this data set below that I want to sort base on mylist from column 'name' as well as acsending by 'A' and descending by 'B'
import pandas as pd
import numpy as np
df1 = pd.DataFrame.from_items([('A', [1, 2, 3]), ('B', [4, 5, 6]) , ('name', ['x','x','x'])])
df2 = pd.DataFrame.from_items([('B', [5, 6, 7]), ('A', [8, 9, 10]) , ('name', ['y','y','y'])])
df3 = pd.DataFrame.from_items([('C', [5, 6, 7]), ('D', [8, 9, 10]), ('A',[1,2,3]), ('B',[4,5,7] ), ('name', ['z','z','z'])])
df_list = [df1,df2,df3[['A','B','name']]]
df = pd.concat(df_list, ignore_index=True)
so my list is
mylist = ['z','x','y']
I want the dataset to start with sort by my list , then sort asc column A then desc column B
is there a way to do this in python ?
======== Edit ==========
I want my final result to be something like
OK, a way to sort by a custom order is to create a dict that defines how 'name' column should be order, call map to add a new column that defines this new order, then call sort and pass in the new column and the others, plus the param ascending where you selectively decide whether each column is sorted ascending or not, and then finally drop that column:
In [20]:
name_sort = {'z':0,'x':1,'y':2}
df['name_sort'] = df.name.map(name_sort)
df
Out[20]:
A B name name_sort
0 1 4 x 1
1 2 5 x 1
2 3 6 x 1
3 8 5 y 2
4 9 6 y 2
5 10 7 y 2
6 1 4 z 0
7 2 5 z 0
8 3 7 z 0
In [23]:
df = df.sort(['name_sort','A','B'], ascending=[1,1,0])
df
Out[23]:
A B name name_sort
6 1 4 z 0
7 2 5 z 0
8 3 7 z 0
0 1 4 x 1
1 2 5 x 1
2 3 6 x 1
3 8 5 y 2
4 9 6 y 2
5 10 7 y 2
In [25]:
df = df.drop('name_sort', axis=1)
df
Out[25]:
A B name
6 1 4 z
7 2 5 z
8 3 7 z
0 1 4 x
1 2 5 x
2 3 6 x
3 8 5 y
4 9 6 y
5 10 7 y
Hi We can do the above issue by using the following:
t = pd.CategoricalDtype(categories=['z','x','y'], ordered=True)
df['sort'] = pd.Series(df.name, dtype=t)
df.sort_values(by=['sort','A','B'], inplace=True)

Categories