CODEWITHSUNDEEP
pythonarrayspandas
lrsppp
lrsppp

Reputation: 1403

Save two arrays (tuples of Date and Value) of different length into pandas array

I have two arrays which consists of tuples (or you can think of 4 arrays).

Date    Array1     Date     Array2
27/10   47         27/10    34
28/10   34         28/10    27
4/11    42         5/11     23
6/11    26

Now I want to store this two arrays into a pandas data frame. Since both arrays have different length, this does not work. The final pandas.DataFrame has to look like this

Date     Array1     Array2
27/10    47         34
28/10    34         27
4/11     42
5/11                23
6/11     26

With empy string or NaN inplace when there is no information for this specific date. Is there any neat way to do that?

Upvotes: 0

Views: 94

Answers (1)

jezrael
jezrael

Reputation: 862691

If input are DataFrames use concat with set_index for align both DataFrames:

df = pd.concat([df1.set_index('Date'), df2.set_index('Date')], axis=1)
print (df)
       Array1  Array2
27/10    47.0    34.0
28/10    34.0    27.0
4/11     42.0     NaN
5/11      NaN    23.0
6/11     26.0     NaN

EDIT:

If there is no year information, pandas add some default - e.g. 1900. It should be important if need ordering by dates like in calendar need to_datetime first:

df1['Date'] = pd.to_datetime(df1['Date'], format='%d/%m')
df2['Date'] = pd.to_datetime(df2['Date'], format='%d/%m')

df = pd.concat([df1.set_index('Date'), df2.set_index('Date')], axis=1).reset_index()

print (df)
        Date  Array1  Array2
0 1900-10-27    47.0    34.0
1 1900-10-28    34.0    27.0
2 1900-11-04    42.0     NaN
3 1900-11-05     NaN    23.0
4 1900-11-06    26.0     NaN

And if need your original format use dt.strftime:

df['Date'] = df['Date'].dt.strftime('%d/%m')
print (df)
    Date  Array1  Array2
0  27/10    47.0    34.0
1  28/10    34.0    27.0
2  04/11    42.0     NaN
3  05/11     NaN    23.0
4  06/11    26.0     NaN

Upvotes: 1

Related Questions