Python: DateTimeIndex not showing in PowerBI

[Anonymous]

Hi, my code runs smoothly in Spyder, so I don't think it's an issue with my code.

However, most of my variables are dataframes, and when I copy the whole code over to the Python Script in Power BI and try to run it, my DatetimeIndex in all my dataframes are no longer shown in the preview, instead it just shows an integer index starting from 1. As a result, when I try to plot my results, it ends up only having one column of data and I can't plot a line graph.

[HALI322]

I got similar issue in getting data from a API datasource (Wind Financial System-A Chinese financial terminal like Bloomberg). Python script code is as follow:

import pandas as pd
from WindPy import *
w.start()

Wind_Data = w.edb("M0039354", "2000-01-01", "2024-11-11","Fill=Previous")
GDP = pd.DataFrame(Wind_Data.Data, columns=Wind_Data.Times, index=Wind_Data.Fields).T
GDP

In Python window, it did feedback transposed dataset include a date index and a column of GDP data. But in Power BI, the return left only GDP data column. The date index column disappeared. Python command line running result below:

>>> import pandas as pd
>>> from WindPy import *
>>> w.start()
.ErrorCode=0
.Data=[Already connected!]
>>>
>>> Wind_Data = w.edb("M0039354", "2000-01-01", "2024-11-11","Fill=Previous")
>>> GDP = pd.DataFrame(Wind_Data.Data, columns=Wind_Data.Times, index=Wind_Data.Fields).T
>>> GDP
                CLOSE
2000-03-31 8.7
2000-06-30 9.1
2000-09-30 8.8
2000-12-31 7.5
2001-03-31 9.5
... ...
2023-09-30 4.9
2023-12-31 5.2
2024-03-31 5.3
2024-06-30 4.7
2024-09-30 4.6

[99 rows x 1 columns]
>>>

Someone help?

(don't worry, all data included are publicly available China GDP growth rate data. No sensitive/private information included.)

[v-yuta-msft]

@Anonymous,

Could share some part of the code? Basically, the interaction between python and power bi should be like pattern below:

#dataset present the current table in power bi as input
#dataset = pd.DataFrame(data, index, columns, dtype, copy)
import pandas as pd

completedData = dataset.fillna(method = 'backfill', inplace = False)
dataset["completedValues"] = completedData[SMI missing values"] #Modify a column

Regards,

Jimmy Tao

[Anonymous]

Hi, part of the code is shown below:

import datetime
import pandas as pd
import numpy as np
import statsmodels.api as sm



def datetimeix(df):
df['Date'] = pd.DatetimeIndex(df['Date'])
df.set_index('Date', inplace = True)
return(df)



df = pd.read_csv(r'C:\Users\blai\Documents\5 year dataset.csv', encoding = "ISO-8859-1") 

mthnsr = cumtotal(df) 

mthnsr = datetimeix(mthnsr)

[v-yuta-msft · Accepted solution]

@Anonymous,

I can reproduce your issue, just add a statement in the button of your code like pattern below:

# 'dataset' holds the input data for this script
import datetime
import pandas as pd
import numpy as np

def datetimeix(df):
    df['Date\t'] = pd.DatetimeIndex(df['Date\t'])
    df.set_index('Date\t', inplace = True)
    return df

df = pd.read_csv(r'C:\Users\JimmyTao\Desktop\Test.csv') 

dataset = df

 

After expand, the result is like this:

 

Community Support Team _ Jimmy Tao

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.