From Wide to Long¶

Tabelo lahko enostavno pretvoriš iz excelove tabele v Markdown z:
https://thisdavej.com/copy-table-in-excel-and-paste-as-a-markdown-table/

WIDE Tabela:

ID	TimeStamp	A	B	C
1	12.05.2019 11:34	0	3	5
2	12.05.2019 12:34	1	1	3
3	12.05.2019 13:34	0	0	3

LONG Tabela:

ID	TimeStamp	Atribut	Vrednost
1	12.05.2019 11:34	A	0
1	12.05.2019 11:34	B	3
1	12.05.2019 11:34	C	5
2	12.05.2019 12:34	A	1
2	12.05.2019 12:34	B	1
2	12.05.2019 12:34	C	3
3	12.05.2019 13:34	A	0
3	12.05.2019 13:34	B	0
3	12.05.2019 13:34	C	3

Python verzija predvorbe iz Wide v Long¶

#podatke so kopiram kar iz excelove tabele
import pandas as pd
df_w = pd.read_clipboard()
print(df_w)

   ID         TimeStamp  A  B  C
0   1  12.05.2019 11:34  0  3  5
1   2  12.05.2019 12:34  1  1  3
2   3  12.05.2019 13:34  0  0  3

#lahko bi naredil df_w kar tukaj:
df_w = pd.DataFrame({
    'ID': [1, 2, 3],
    'TimeStamp': ['12.05.2019 11:34', '12.05.2019 12:34', '12.05.2019 13:34'],
    'A': [0, 1, 0],
    'B': [3, 1, 0],
    'C': [5, 3, 3]
})
print(pdf_w)

   ID         TimeStamp  A  B  C
0   1  12.05.2019 11:34  0  3  5
1   2  12.05.2019 12:34  1  1  3
2   3  12.05.2019 13:34  0  0  3

df_w = df_w.set_index(['ID', 'TimeStamp'])
print(df_w)

                     A  B  C
ID TimeStamp                
1  12.05.2019 11:34  0  3  5
2  12.05.2019 12:34  1  1  3
3  12.05.2019 13:34  0  0  3

df_w = df_w.reset_index()
print(df_w)

   ID         TimeStamp  A  B  C
0   1  12.05.2019 11:34  0  3  5
1   2  12.05.2019 12:34  1  1  3
2   3  12.05.2019 13:34  0  0  3

#TO JE PRETVORBA iz WIDE v LONG
df_l = pd.melt(df_w, id_vars = ['ID', 'TimeStamp'])   # pd.melt(df_w, id_vars = ['ID', 'TimeStamp'], value_vars = ['A','B','C'])
print(df_l)

   ID         TimeStamp variable  value
0   1  12.05.2019 11:34        A      0
1   2  12.05.2019 12:34        A      1
2   3  12.05.2019 13:34        A      0
3   1  12.05.2019 11:34        B      3
4   2  12.05.2019 12:34        B      1
5   3  12.05.2019 13:34        B      0
6   1  12.05.2019 11:34        C      5
7   2  12.05.2019 12:34        C      3
8   3  12.05.2019 13:34        C      3

nazaj v excel kopiram z:

df_l.to_clipboard(excel=True)

#Za lepsi prikaz
from IPython.display import display, HTML
HTML(df_l.to_html())