I am using groupby-apply to create new DataFrame from given Data Frame. But if given DataFrame is empty result would look like given DataFrame with group keys not like target new DataFrame. So to get look of target new DataFrame I have to use if..else with length check and if given DataFrame is empty then manually create DataFrame with specified columns and indexes.
It is kinda broken flow of code. Also if in future structure of target DataFrame happen to change I would have to fix code in two places instead of one.
Is there a way to get look of a target DataFrame even if given DataFrame is empty with GroupBy only (or without if..else)?
Simplified example:
def some_func(df: pd.DataFrame):
return df.values.sum() + pd.DataFrame([[1,1,1], [2,2,2], [3,3,3]], columns=['new_col1', 'new_col2', 'new_col3'])
df1 = pd.DataFrame([[1,1], [1,2], [2,1], [2,2]], columns=['col1', 'col2'])
df2 = pd.DataFrame(columns=['col1', 'col2'])
df1_grouped = df1.groupby(['col1'], group_keys=False).apply(lambda df: some_func(df))
df2_grouped = df2.groupby(['col1'], group_keys=False).apply(lambda df: some_func(df))
Result for df1 is ok:
new_col1 new_col2 new_col3
0 6 6 6
1 7 7 7
2 8 8 8
0 8 8 8
1 9 9 9
2 10 10 10
And not ok for df2:
Empty DataFrame
Columns: [col1, col2]
Index: []
If..else to get expected result for df2:
df = df2
if df.empty:
df_grouped = pd.DataFrame(columns=['new_col1', 'new_col2', 'new_col3'])
else:
df_grouped = df.groupby(['col1'], group_keys=False).apply(lambda df: some_func(df))
Gives what I need:
Empty DataFrame
Columns: [new_col1, new_col2, new_col3]
Index: []
Related
There are a total of 4 dataframes (df1 / df2 / df3 / df4),
Each dataframe has a different template, but they all have the same columns.
I want to merges the row of each dataframe based on the same column, but what function should I use? A 'merge' or 'join' function doesn't seem to work, and deleting the rest of the columns after grouping them into a list seems to be too messy.
I want to make attached image
This is an option, you can merge the dataframes and then drop the useless columns from the total dataframe.
df_total = pd.concat([df1, df2, df3, df4], axis=0)
df_total.drop(['Value2', 'Value3'], axis=1)
You can use reduce to get it done too.
from functools import reduce
reduce(lambda left,right: pd.merge(left, right, on=['ID','value1'], how='outer'), [df1,df2,df3,df4])[['ID','value1']]
ID value1
0 a 1
1 b 4
2 c 5
3 f 1
4 g 5
5 h 6
6 i 1
I have two data frames as shown below df1 and df2. I want to create a third dataframe i.e. df as shown below. What would be the appropriate way?
df1={'id':['a','b','c'],
'val':[1,2,3]}
df1=pd.DataFrame(df)
df1
id val
0 a 1
1 b 2
2 c 3
df2={'yr':['2010','2011','2012'],
'val':[4,5,6]}
df2=pd.DataFrame(df2)
df2
yr val
0 2010 4
1 2011 5
2 2012 6
df={'id':['a','b','c'],
'val':[1,2,3],
'2010':[4,8,12],
'2011':[5,10,15],
'2012':[6,12,18]}
df=pd.DataFrame(df)
df
id val 2010 2011 2012
0 a 1 4 5 6
1 b 2 8 10 12
2 c 3 12 15 18
I can basically convert df1 and df2 as 1 by n matrices and get n by n result and assign it back to the df1. But is there any easy pandas way?
TL;DR
We can do it in one line like this:
df1.join(df1.val.apply(lambda x: x * df2.set_index('yr').val))
or like this:
df1.join(df1.set_index('id') # df2.set_index('yr').T, on='id')
Done.
The long story
Let's see what's going on here.
To find the output of multiplication of each df1.val by values in df2.val we use apply:
df1['val'].apply(lambda x: x * df2.val)
The function inside will obtain df1.vals one by one and multiply each by df2.val element-wise (see broadcasting for details if needed). As far as df2.val is a pandas sequence, the output is a data frame with indexes df1.val.index and columns df2.val.index. By df2.set_index('yr') we force years to be indexes before multiplication so they will become column names in the output.
DataFrame.join is joining frames index-on-index by default. So due to identical indexes of df1 and the multiplication output, we can apply df1.join( <the output of multiplication> ) as is.
At the end we get the desired matrix with indexes df1.index and columns id, val, *df2['yr'].
The second variant with # operator is actually the same. The main difference is that we multiply 2-dimentional frames instead of series. These are the vertical and horizontal vectors, respectively. So the matrix multiplication will produce a frame with indexes df1.id and columns df2.yr and element-wise multiplication as values. At the end we connect df1 with the output on identical id column and index respectively.
This works for me:
df2 = df2.T
new_df = pd.DataFrame(np.outer(df1['val'],df2.iloc[1:]))
df = pd.concat([df1, new_df], axis=1)
df.columns = ['id', 'val', '2010', '2011', '2012']
df
The output I get:
id val 2010 2011 2012
0 a 1 4 5 6
1 b 2 8 10 12
2 c 3 12 15 18
Your question is a bit vague. But I suppose you want to do something like that:
df = pd.concat([df1, df2], axis=1)
I have the following
df1 = pd.DataFrame({'data': [1,2,3]})
df2 = pd.DataFrame({'data': [4,5,6]})
df = pd.concat([df1,df2], keys=['hello','world'], axis=1)
What is the "proper" way of creating a new nested column (say, df['world']['data']*2) within the hello column? I have tried df['hello']['new_col'] = df['world']['data']*2 but this does not seem to work.
Use tuples for select and set MultiIndex:
df[('hello','new_col')] = df[('world','data')]*2
print (df)
hello world hello
data data new_col
0 1 4 8
1 2 5 10
2 3 6 12
Selecting like df['world']['data'] is not recommended - link, because possible chained indexing.
I'm a biologist and very new to Python (I use v3.5) and pandas. I have a pandas dataframe (df), from which I need to make several dataframes (df1... dfn) that can be placed in a dictionary (dictA), which currently has the correct number (n) of empty dataframes. I also have a dictionary (dictB) of n (individual) lists of column names that were extracted from df. The keys in 2 dictionaries match. I'm trying to append the empty dfs within dictA with parts of df based on the column names within the lists in dictB.
import pandas as pd
listA=['A', 'B', 'C',...]
dictA={i:pd.DataFrame() for i in listA}
lets say I have something like this:
dictA={'A': df1, 'B': df2}
dictB={'A': ['A1', A2', 'A3'],
'B': ['B1', B2']}
df=pd.DataFrame({'A1': [0,2,4,5],
'A2': [2,5,6,7],
'A3': [5,6,7,8],
'B1': [2,5,6,7],
'B2': [1,3,5,6]})
listA=['A', 'B']
what I'm trying to get is for df1 and df2 to get appended with portions of df like this, so that the output for df1 is like this:
A1 A2 A3
0 0 2 5
1 2 4 6
2 4 6 7
3 5 7 8
df2 would have columns B1 and B2.
I tried the following loop and some alterations, but it doesn't yield populated dfs:
for key, values in dictA.items():
values.append(df[dictB[key]])
Thanks and sorry if this was already addressed elsewhere but I couldn't find it.
You could create the dataframes you want like this instead :
df = #Your original dataframe containing all the columns
df_A = df.iloc[:][[col for col in df if 'A' in col]]
df_B = df.iloc[:][[col for col in df if 'B' in col]]
I tried following code to select columns from a dataframe. My dataframe has about 50 values. At the end, I want to create the sum of selected columns, create a new column with these sum values and then delete the selected columns.
I started with
columns_selected = ['A','B','C','D','E']
df = df[df.column.isin(columns_selected)]
but it said AttributeError: 'DataFrame' object has no attribute 'column'
Regarding the sum: As I don't want to write for the sum
df['sum_1'] = df['A']+df['B']+df['C']+df['D']+df['E']
I also thought that something like
df['sum_1'] = df[columns_selected].sum(axis=1)
would be more convenient.
You want df[columns_selected] to sub-select the df by a list of columns
you can then do df['sum_1'] = df[columns_selected].sum(axis=1)
To filter the df to just the cols of interest pass a list of the columns, df = df[columns_selected] note that it's a common error to just a list of strings: df = df['a','b','c'] which will raise a KeyError.
Note that you had a typo in your original attempt:
df = df.loc[:,df.columns.isin(columns_selected)]
The above would've worked, firstly you needed columns not column, secondly you can use the boolean mask as a mask against the columns by passing to loc or ix as the column selection arg:
In [49]:
df = pd.DataFrame(np.random.randn(5,5), columns=list('abcde'))
df
Out[49]:
a b c d e
0 -0.778207 0.480142 0.537778 -1.889803 -0.851594
1 2.095032 1.121238 1.076626 -0.476918 -0.282883
2 0.974032 0.595543 -0.628023 0.491030 0.171819
3 0.983545 -0.870126 1.100803 0.139678 0.919193
4 -1.854717 -2.151808 1.124028 0.581945 -0.412732
In [50]:
cols = ['a','b','c']
df.ix[:, df.columns.isin(cols)]
Out[50]:
a b c
0 -0.778207 0.480142 0.537778
1 2.095032 1.121238 1.076626
2 0.974032 0.595543 -0.628023
3 0.983545 -0.870126 1.100803
4 -1.854717 -2.151808 1.124028