Pairwise matrix counts of two columns using pandas [duplicate] - python

This question already has answers here:
How can I pivot a dataframe?
(5 answers)
Closed 2 years ago.
I am trying to obtain pairwise counts of two column variables using pandas. I have a dataframe of two columns in the following format:
col1 col2
a e
b g
c h
d f
a g
b h
c f
d e
a f
b g
c g
d h
a e
b e
c g
d h
b h
What I would like to get as output would be the following matrix of counts, for e.g.:
e f g h
a 2 1 1 0
b 1 0 2 2
c 0 1 2 1
d 1 1 0 2
I am getting totally confused with pandas iterating over columns, rows, indexes and such. Appreciate some guidance here.

Pandas often has simple functions built in - in this case, you want crosstab:
pd.crosstab(dat['col1'], dat['col2'])
full code:
import pandas as pd
from io import StringIO
x = '''col1 col2
a e
b g
c h
d f
a g
b h
c f
d e
a f
b g
c g
d h
a e
b e
c g
d h
b h'''
dat = pd.read_csv(StringIO(x), sep = '\s+')
pd.crosstab(dat['col1'], dat['col2'])

You're looking for a crosstab:
count_matrix = pd.crosstab(index=df["col1"], columns=df["col2"])
print(count_matrix)
col2 e f g h
col1
a 2 1 1 0
b 1 0 2 2
c 0 1 2 1
d 1 1 0 2
If you don't like the column/index names in (e.g. still seeing "col1" and "col2"), then you can remove them with rename_axis:
count_matrix = count_matrix.rename_axis(index=None, columns=None)
print(count_matrix)
e f g h
a 2 1 1 0
b 1 0 2 2
c 0 1 2 1
d 1 1 0 2
If you want that all together in one snippet:
count_matrix = (pd.crosstab(index=df["col1"], columns=df["col2"])
.rename_axis(index=None, columns=None))

Related

How to convert binary columns with multiple occurrences into categorical data in Pandas

I have the following example data set
A
B
C
D
foo
0
1
1
bar
0
0
1
baz
1
1
0
How could extract the column names of each 1 occurrence in a row and put that into another column E so that I get the following table:
A
B
C
D
E
foo
0
1
1
C, D
bar
0
0
1
D
baz
1
1
0
B, C
Note that there can be more than two 1s per row.
You can use DataFrame.dot.
df['E'] = df[['B', 'C', 'D']].dot(df.columns[1:] + ', ').str.rstrip(', ')
df
A B C D E
0 foo 0 1 1 C, D
1 bar 0 0 1 D
2 baz 1 1 0 B, C
Inspired by jezrael's answer in this post.
Another way is that you can convert each row to boolean and use it as a selection mask to filter the column names.
cols = pd.Index(['B', 'C', 'D'])
df['E'] = df[cols].astype('bool').apply(lambda row: ", ".join(cols[row]), axis=1)
df
A B C D E
0 foo 0 1 1 C, D
1 bar 0 0 1 D
2 baz 1 1 0 B, C

pandas dataframe group columns based on name and apply a function

I have a dataframe:
df = [A B C D E_p0 E_p1 E_p2 K_p0 K_p1 K_2
a 2 r 4 3 6 1 9 5 1
e g 1 d 5 8 2 7 1 4]
And I want to group columns based on the prefix and aggregate them by a function, such as mean or max or rms.
So, for example if my function is max, the output is:
df = [A B C D E K
a 2 r 4 6 9
e g 1 d 8 7 ]
You can convert columns without separator to index and then grouping with lambda function per columns with aggregate function like max:
m = df.columns.str.contains('_')
df = (df.set_index(df.columns[~m].tolist())
.groupby(lambda x: x.split('_')[0], axis=1)
.max()
.reset_index())
print (df)
A B C D E K
0 a 2 r 4 6 9
1 e g 1 d 8 7
Solution with custom function:
def rms(x):
return np.sqrt(np.sum(x**2, axis=1)/len(x.columns))
m = df.columns.str.contains('_')
df1 = (df.set_index(df.columns[~m].tolist())
.groupby(lambda x: x.split('_')[0], axis=1)
.agg(rms)
.reset_index())
print (df1)
A B C D E K
0 a 2 r 4 3.915780 5.972158
1 e g 1 d 5.567764 4.690416

Expand pandas dataframe by replacing cell value with a list

I have a pandas dataframe like this below:
A B C
a b c
d e f
where A B and C are column names. Now i have a list:
mylist = [1,2,3]
I want to replace the c in column C with list such as dataframe expands for all value of list, like below:
A B C
a b 1
a b 2
a b 3
d e f
Any help would be appreciated!
I tried this,
mylist = [1,2,3]
x=pd.DataFrame({'mylist':mylist})
x['C']='c'
res= pd.merge(df,x,on=['C'],how='left')
res['mylist']=res['mylist'].fillna(res['C'])
For further,
del res['C']
res.rename(columns={"mylist":"C"},inplace=True)
print res
Output:
A B C
0 a b 1
1 a b 2
2 a b 3
3 d e f
You can use:
print (df)
A B C
0 a b c
1 d e f
2 a b c
3 t e w
mylist = [1,2,3]
idx1 = df.index[df.C == 'c']
df = df.loc[idx1.repeat(len(mylist))].assign(C=mylist * len(idx1)).append(df[df.C != 'c'])
print (df)
A B C
0 a b 1
0 a b 2
0 a b 3
2 a b 1
2 a b 2
2 a b 3
1 d e f
3 t e w

Selecting columns from a pandas dataframe based on columns conditions

I want to select to new dataframe, columns that have 'C' in value
protein 1 2 3 4 5
prot1 C M D F A
prot2 C D A M A
prot3 C C D F A
prot4 S D F C L
prot5 S D A I L
So i want to have this:
protein 1 2 4
prot1 C M F
prot2 C D M
prot3 C C F
prot4 S D C
prot5 S D I
Number of colums can be n, i found examples only which i must specify column name... i cant do this here. The script should check column by colummn.
In [22]: df[['protein']].join(df[df.columns[df.eq('C').any()]])
Out[22]:
protein 1 2 4
0 prot1 C M F
1 prot2 C D M
2 prot3 C C F
3 prot4 S D C
4 prot5 S D I
Use:
np.random.seed(123)
n = np.random.choice(['C','M','D', '-'], size=(3,10))
n[:,0] = ['a','b','w']
foo = pd.DataFrame(n)
print (foo)
0 1 2 3 4 5 6 7 8 9
0 a M D D C D D M - D
1 b M D M C M D - M C
2 w C - M - D M C C C
mask = foo.eq('C').any()
#set columns which need in output
mask.loc[0] = True
#filter
print (foo.loc[:,mask])
0 1 4 7 8 9
0 a M C M - D
1 b M C - M C
2 w C - C C C

Convert Two column data frame to occurrence matrix in pandas

Hi all I have a csv file which contains data as the format below
A a
A b
B f
B g
B e
B h
C d
C e
C f
The first column contains items second column contains available feature from feature vector=[a,b,c,d,e,f,g,h]
I want to convert this to occurence matrix look like below
a,b,c,d,e,f,g,h
A 1,1,0,0,0,0,0,0
B 0,0,0,0,1,1,1,1
C 0,0,0,1,1,1,0,0
Can anyone tell me how to do this using pandas?
Here is another way to do it using pd.get_dummies().
import pandas as pd
# your data
# =======================
df
col1 col2
0 A a
1 A b
2 B f
3 B g
4 B e
5 B h
6 C d
7 C e
8 C f
# processing
# ===================================
pd.get_dummies(df.col2).groupby(df.col1).apply(max)
a b d e f g h
col1
A 1 1 0 0 0 0 0
B 0 0 0 1 1 1 1
C 0 0 1 1 1 0 0
Unclear if your data has a typo or not but you can crosstab for this:
In [95]:
pd.crosstab(index=df['A'], columns = df['a'])
Out[95]:
a b d e f g h
A
A 1 0 0 0 0 0
B 0 0 1 1 1 1
C 0 1 1 1 0 0
In your sample data your second column has value a as the name of that column but in your expected output it's in the column as a value
EDIT
OK I fixed your input data so it generates the correct result:
In [98]:
import pandas as pd
import io
t="""A a
A b
B f
B g
B e
B h
C d
C e
C f"""
df = pd.read_csv(io.StringIO(t), sep='\s+', header=None, names=['A','a'])
df
Out[98]:
A a
0 A a
1 A b
2 B f
3 B g
4 B e
5 B h
6 C d
7 C e
8 C f
In [99]:
ct = pd.crosstab(index=df['A'], columns = df['a'])
ct
Out[99]:
a a b d e f g h
A
A 1 1 0 0 0 0 0
B 0 0 0 1 1 1 1
C 0 0 1 1 1 0 0
This approach yields the same result in a scipy sparse coo matrix much faster
from scipy import sparse
df['col1'] = df['col1'].astype("category")
df['col2'] = df['col2'].astype("category")
df['ones'] = 1
user_items = sparse.coo_matrix((df.ones.astype(float),
(df.col1.cat.codes,
df.col2.cat.codes)))

Categories

Resources