[Python] csv file wide to long. Float value gets truncated and getting integer for index column

I have csv file with approx 40K rows and 1200 columns in wide format. with looks something like this:

| Millisec_diff | Value1 | Value2 | Value3 |
| ------------- | ------ | ------ |------- |
| 0 | 100 | 200 | 1.3 |
| 0.005 | 101 | 20.1 | 1.3 |
| 0.01 | 103 | 20.1 | 2.5 |
| 0.015 | 104 | 24.1 | 4.5 |
| 0.02 | 103 | 40.1 | 5.4 |


I added date time column based Millisec_diff and converted to long format using pandas.melt:

start = datetime.today()
df['time'] = (
datetime(
start.year,
start.month,
start.day,
start.hour,
start.minute,
start.second,
start.microsecond)
+ pandas.to_timedelta(df['Millisec_diff '], unit='ms')
)

df_long = pandas.melt(
df,
id_vars=["time", 'Millisec_diff '],
value_vars=col_names,
var_name='param',
value_name='value')


When I run the code on sample dataset, I can see proper values (eg: .005, .01) for Millisec_diff but when I run this on whole file, I can see only integers (like 1, 2) for Millisec_diff instead of actual decimal values. I am not sure what is happening.

Continue reading...