I need to create groups using two columns. For example, I took shop_id and week. Here is the df:
shop_id week
0 1 1
1 1 2
2 1 3
3 2 1
4 2 2
5 3 2
6 1 5
Imagine that each group is some promo which took place in each shop consecutively (week by week). So, my attempt was to use sorting, shifting by 1 to get last_week, use booleans and then iterate over them, incrementing each time whereas condition not met:
test_df = pd.DataFrame({'shop_id':[1,1,1,2,2,3,1], 'week':[1,2,3,1,2,2,5]})
def createGroups(df, shop_id, week, group):
'''Create groups where is the same shop_id and consecutive week
'''
periods = []
period = 0
# sorting to create chronological order
df = df.sort_values(by = [shop_id,week],ignore_index = True)
last_week = df[week].shift(+1)==df[week]-1
last_shop = df[shop_id].shift(+1)==df[shop_id]
# here i iterate over booleans and increment group by 1
# if shop is different or last period is more than 1 week ago
for p,s in zip(last_week,last_shop):
if (p == True) and (s == True):
periods.append(period)
else:
period += 1
periods.append(period)
df[group] = periods
return df
createGroups(test_df, 'shop_id', 'week', 'promo')
And I get the grouping I need:
shop_id week promo
0 1 1 1
1 1 2 1
2 1 3 1
3 1 5 2
4 2 1 3
5 2 2 3
6 3 2 4
However, function seems to be an overkill. Any ideas on how to get the same without a for-loop using native pandas function? I saw .ngroups() in docs but have no idea how to apply it to my case. Even better would be to vectorise it somehow, but I don't know how to achieve this:(
First we want to identify the promotions (continuously in weeks), then use groupby().ngroup() to enumerate the promotion:
df = df.sort_values('shop_id')
s = df['week'].diff().ne(1).groupby(df['shop_id']).cumsum()
df['promo'] = df.groupby(['shop_id',s]).ngroup() + 1
Update: This is based on your solution:
df = df.sort_values(['shop_id','week'])
s = df[['shop_id', 'week']]
df['promo'] = (s['shop_id'].ne(s['shop_id'].shift()) |
s['week'].diff().ne(1) ).cumsum()
Output:
shop_id week promo
0 1 1 1
1 1 2 1
2 1 3 1
6 1 5 2
3 2 1 3
4 2 2 3
5 3 2 4
Related
I have a big dataframe with more than 100 columns. I am sharing a miniature version of my real dataframe below
ID rev_Q1 rev_Q5 rev_Q4 rev_Q3 rev_Q2 tx_Q3 tx_Q5 tx_Q2 tx_Q1 tx_Q4
1 1 1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
I would like to do the below
a) sort the column names based on Quarters (ex:Q1,Q2,Q3,Q4,Q5..Q100..Q1000) for each column pattern
b) By column pattern, I mean the keyword that is before underscore which is rev and tx.
So, I tried the below but it doesn't work and it also shifts the ID column to the back
df = df.reindex(sorted(df.columns), axis=1)
I expect my output to be like as below. In real time, there are more than 100 columns with more than 30 patterns like rev, tx etc. I want my ID column to be in the first position as shown below.
ID rev_Q1 rev_Q2 rev_Q3 rev_Q4 rev_Q5 tx_Q1 tx_Q2 tx_Q3 tx_Q4 tx_Q5
1 1 1 1 1 1 1 1 1 1 1
2 1 1 1 1 1 1 1 1 1 1
For the provided example, df.sort_index(axis=1) should work fine.
If you have Q values higher that 9, use natural sorting with natsort:
from natsort import natsort_key
out = df.sort_index(axis=1, key=natsort_key)
Or using manual sorting with np.lexsort:
idx = df.columns.str.split('_Q', expand=True, n=1)
order = np.lexsort([idx.get_level_values(1).astype(float), idx.get_level_values(0)])
out = df.iloc[:, order]
Something like:
new_order = list(df.columns)
new_order = ['ID'] + sorted(new_order.remove("ID"))
df = df[new_order]
we manually put "ID" in front and then sort what is remaining
The idea is to create a dataframe from the column names. Create two columns: one for Variable and another one for Quarter number. Finally sort this dataframe by values then extract index.
idx = (df.columns.str.extract(r'(?P<V>[^_]+)_Q(?P<Q>\d+)')
.fillna(0).astype({'Q': int})
.sort_values(by=['V', 'Q']).index)
df = df.iloc[:, idx]
Output:
>>> df
ID rev_Q1 rev_Q2 rev_Q3 rev_Q4 rev_Q5 tx_Q1 tx_Q2 tx_Q3 tx_Q4 tx_Q5
0 1 1 1 1 1 1 1 1 1 1 1
1 2 1 1 1 1 1 1 1 1 1 1
>>> (df.columns.str.extract(r'(?P<V>[^_]+)_Q(?P<Q>\d+)')
.fillna(0).astype({'Q': int})
.sort_values(by=['V', 'Q']))
V Q
0 0 0
1 rev 1
5 rev 2
4 rev 3
3 rev 4
2 rev 5
9 tx 1
8 tx 2
6 tx 3
10 tx 4
7 tx 5
so i have this code:
import pandas as pd
id_1=[0,0,0,0,0,0,2,0,4,5,6,7,1,0,5,3]
exp_1=[1,2,3,4,5,6,1,7,1,1,1,1,1,8,2,1]
df = pd.DataFrame(list(zip(id_1,exp_1)), columns =['Patch', 'Exploit'])
df = (
df.groupby((df.Patch != df.Patch.shift(1)).cumsum())
.agg({"Patch": ("first", "count")})
.reset_index(drop=True)
)
print(df)
the output is:
Patch
first count
0 0 6
1 2 1
2 0 1
3 4 1
4 5 1
5 6 1
6 7 1
7 1 1
8 0 1
9 5 1
10 3 1
I wanted to create a data frame with a new column called count where I can store the consecutive appearance of the patch (id_1).
However, the above code creates a dictionary of the patch and I don't know how to individually manipulate only the values stored in the column called count.
suppose I want to remove all the 0 from id_1 and then count the consecutive appearance.
or I have to find the average of the count column only then?
If you want to remove all 0 from column Patch, then you can filter the dataframe just before .groupby. For example:
df = (
df[df.Patch != 0]
.groupby((df.Patch != df.Patch.shift(1)).cumsum())
.agg({"Patch": ("first", "count")})
.reset_index(drop=True)
)
print(df)
Prints:
Patch
first count
0 2 1
1 4 1
2 5 1
3 6 1
4 7 1
5 1 1
6 5 1
7 3 1
Let's assume the input dataset:
test1 = [[0,7,50], [0,3,51], [0,3,45], [1,5,50],[1,0,50],[2,6,50]]
df_test = pd.DataFrame(test1, columns=['A','B','C'])
that corresponds to:
A B C
0 0 7 50
1 0 3 51
2 0 3 45
3 1 5 50
4 1 0 50
5 2 6 50
I would like to obtain the a dataset grouped by 'A', together with the most common value for 'B' in each group, and the occurrences of that value:
A most_freq freq
0 3 2
1 5 1
2 6 1
I can obtain the first 2 columns with:
grouped = df_test.groupby("A")
out_df = pd.DataFrame(index=grouped.groups.keys())
out_df['most_freq'] = df_test.groupby('A')['B'].apply(lambda x: x.value_counts().idxmax())
but I am having problems the last column.
Also: is there a faster way that doesn't involve 'apply'? This solution doesn't scale well with lager inputs (I also tried dask).
Thanks a lot!
Use SeriesGroupBy.value_counts which sorting by default, so then add DataFrame.drop_duplicates for top values after Series.reset_index:
df = (df_test.groupby('A')['B']
.value_counts()
.rename_axis(['A','most_freq'])
.reset_index(name='freq')
.drop_duplicates('A'))
print (df)
A most_freq freq
0 0 3 2
2 1 0 1
4 2 6 1
Suppose I have pandas DataFrame like this:
df = pd.DataFrame({'id':[1,1,1,2,2,2,2,3,4], 'value':[1,2,3,1,2,3,4,1,1]})
which looks like:
id value
0 1 1
1 1 2
2 1 3
3 2 1
4 2 2
5 2 3
6 2 4
7 3 1
8 4 1
I want to get a new DataFrame with top 2 records for each id, like this:
id value
0 1 1
1 1 2
3 2 1
4 2 2
7 3 1
8 4 1
I can do it with numbering records within group after groupby:
dfN = df.groupby('id').apply(lambda x:x['value'].reset_index()).reset_index()
which looks like:
id level_1 index value
0 1 0 0 1
1 1 1 1 2
2 1 2 2 3
3 2 0 3 1
4 2 1 4 2
5 2 2 5 3
6 2 3 6 4
7 3 0 7 1
8 4 0 8 1
then for the desired output:
dfN[dfN['level_1'] <= 1][['id', 'value']]
Output:
id value
0 1 1
1 1 2
3 2 1
4 2 2
7 3 1
8 4 1
But is there more effective/elegant approach to do this? And also is there more elegant approach to number records within each group (like SQL window function row_number()).
Did you try
df.groupby('id').head(2)
Output generated:
id value
id
1 0 1 1
1 1 2
2 3 2 1
4 2 2
3 7 3 1
4 8 4 1
(Keep in mind that you might need to order/sort before, depending on your data)
EDIT: As mentioned by the questioner, use
df.groupby('id').head(2).reset_index(drop=True)
to remove the MultiIndex and flatten the results:
id value
0 1 1
1 1 2
2 2 1
3 2 2
4 3 1
5 4 1
Since 0.14.1, you can now do nlargest and nsmallest on a groupby object:
In [23]: df.groupby('id')['value'].nlargest(2)
Out[23]:
id
1 2 3
1 2
2 6 4
5 3
3 7 1
4 8 1
dtype: int64
There's a slight weirdness that you get the original index in there as well, but this might be really useful depending on what your original index was.
If you're not interested in it, you can do .reset_index(level=1, drop=True) to get rid of it altogether.
(Note: From 0.17.1 you'll be able to do this on a DataFrameGroupBy too but for now it only works with Series and SeriesGroupBy.)
Sometimes sorting the whole data ahead is very time consuming.
We can groupby first and doing topk for each group:
g = df.groupby(['id']).apply(lambda x: x.nlargest(topk,['value'])).reset_index(drop=True)
df.groupby('id').apply(lambda x : x.sort_values(by = 'value', ascending = False).head(2).reset_index(drop = True))
Here sort values ascending false gives similar to nlargest and True gives similar to nsmallest.
The value inside the head is the same as the value we give inside nlargest to get the number of values to display for each group.
reset_index is optional and not necessary.
This works for duplicated values
If you have duplicated values in top-n values, and want only unique values, you can do like this:
import pandas as pd
ifile = "https://raw.githubusercontent.com/bhishanpdl/Shared/master/data/twitter_employee.tsv"
df = pd.read_csv(ifile,delimiter='\t')
print(df.query("department == 'Audit'")[['id','first_name','last_name','department','salary']])
id first_name last_name department salary
24 12 Shandler Bing Audit 110000
25 14 Jason Tom Audit 100000
26 16 Celine Anston Audit 100000
27 15 Michale Jackson Audit 70000
If we do not remove duplicates, for the audit department we get top 3 salaries as 110k,100k and 100k.
If we want to have not-duplicated salaries per each department, we can do this:
(df.groupby('department')['salary']
.apply(lambda ser: ser.drop_duplicates().nlargest(3))
.droplevel(level=1)
.sort_index()
.reset_index()
)
This gives
department salary
0 Audit 110000
1 Audit 100000
2 Audit 70000
3 Management 250000
4 Management 200000
5 Management 150000
6 Sales 220000
7 Sales 200000
8 Sales 150000
To get the first N rows of each group, another way is via groupby().nth[:N]. The outcome of this call is the same as groupby().head(N). For example, for the top-2 rows for each id, call:
N = 2
df1 = df.groupby('id', as_index=False).nth[:N]
To get the largest N values of each group, I suggest two approaches.
First sort by "id" and "value" (make sure to sort "id" in ascending order and "value" in descending order by using the ascending parameter appropriately) and then call groupby().nth[].
N = 2
df1 = df.sort_values(by=['id', 'value'], ascending=[True, False])
df1 = df1.groupby('id', as_index=False).nth[:N]
Another approach is to rank the values of each group and filter using these ranks.
# for the entire rows
N = 2
msk = df.groupby('id')['value'].rank(method='first', ascending=False) <= N
df1 = df[msk]
# for specific column rows
df1 = df.loc[msk, 'value']
Both of these are much faster than groupby().apply() and groupby().nlargest() calls as suggested in the other answers on here(1, 2, 3). On a sample with 100k rows and 8000 groups, a %timeit test showed that it was 24-150 times faster than those solutions.
Also, instead of slicing, you can also pass a list/tuple/range to a .nth() call:
df.groupby('id', as_index=False).nth([0,1])
# doesn't even have to be consecutive
# the following returns 1st and 3rd row of each id
df.groupby('id', as_index=False).nth([0,2])
I have a Pandas script that counts the number of readmissions to hospital within 30 days based on a few conditions. I wonder if it could be vectorized to improve performance. I've experimented with df.rolling().apply, but so far without luck.
Here's a table with contrived data to illustrate:
ID VISIT_NO ARRIVED LEFT HAD_A_MASSAGE BROUGHT_A_FRIEND
1 1 29/02/1996 01/03/1996 0 1
1 2 01/12/1996 04/12/1996 1 0
2 1 20/09/1996 21/09/1996 1 0
3 1 27/06/1996 28/06/1996 1 0
3 2 04/07/1996 06/07/1996 0 1
3 3 16/07/1996 18/07/1996 0 1
4 1 21/02/1996 23/02/1996 0 1
4 2 29/04/1996 30/04/1996 1 0
4 3 02/05/1996 02/05/1996 0 1
4 4 02/05/1996 03/05/1996 0 1
5 1 03/10/1996 05/10/1996 1 0
5 2 07/10/1996 08/10/1996 0 1
5 3 10/10/1996 11/10/1996 0 1
First, I create a dictionary with IDs:
ids = massage_df[massage_df['HAD_A_MASSAGE'] == 1]['ID']
id_dict = {id:0 for id in ids}
Everybody in this table has had a massage, but in my real dataset, not all people are so lucky.
Next, I run this bit of code:
for grp, df in massage_df.groupby(['ID']):
date_from = df.loc[df[df['HAD_A_MASSAGE']==1].index, 'LEFT']
date_to = date_from + DateOffset(days=30)
mask = ((date_from.values[0] < df['ARRIVED']) &
(df['ARRIVED'] <= date_to.values[0]) &
(df['BROGHT_A_FRIEND'] == 1))
if len(df[mask]) > 0:
id_dict[df['ID'].iloc[0]] = len(df[mask])
Basically, I want to count the number of times when someone originally came in for a massage (single or with a friend) and then came back within 30 days with a friend. The expected results for this table would be a total of 6 readmissions for IDs 3, 4 and 5.