Can't we append a large number of time series together?

Asked Mar 03 '18 at 19:14

Active Mar 03 '18 at 19:25

Viewed 28 times

I am facing silly error with my append operation: suppose I have about 600 separate dataframes of shape (6,10) and I want to concatenate them in one big dataframe with append() in a for loop. Doing this for the first 100 df works just fine but it generates the following error once I reach 300 dataframes:

"Reindexing only valid with uniquely valued Index objects"

here is the code snippet:

for i in range(400):
## Do some processing
temp =  pd.concat([temp,ts.iloc[:,1]], axis=1)
df_pred_ult = df_pred_ult.append(temp)

So , my question, doesn't append() work well for a huge number of df? Is there a better way to do so...

edited Mar 03 '18 at 19:25

asked Mar 03 '18 at 19:14

Galileo

1

What about `df = pd.concat([df1, df2, df3, df4, df5, dfn], ignore_index=True)` ? – jezrael Mar 03 '18 at 19:21
Possible duplicate - [link](https://stackoverflow.com/q/36526282/2901002) – jezrael Mar 03 '18 at 19:22
df_pred_ult = pd.concat([df_pred_ult,temp], axis=0, ignore_index=True) – Galileo Mar 03 '18 at 19:26
No duplicate rows , I have already checked... – Galileo Mar 03 '18 at 19:27
1

Is possible create list of all dataframes and then use `concat` ? – jezrael Mar 03 '18 at 19:28

Can't we append a large number of time series together?

0 Answers0