Finding the difference in time

[Van Ha]

HI All:
I am unable to replicate a variable that I was able to previously generate. I am trying to finding the difference between time expressed in minutes using the command:
generate onset_hospitarrival = Clockdiff(hosp_datetime , str_datetime, "minute")

I wonder if it is a formatting issue because I don't see the below values on my screen but I tried %tC, %tc, and %td with no success. Thanks in advance!

* Example generated by -dataex-. For more info, type help dataex
clear
input double(str_datetime hosp_datetime) float onset_hospitarrival
1.8099516e+12 1.8099597e+12 -135
1.8009639e+12  1.800972e+12 -135
1.8008136e+12 1.8008148e+12  -20
1.7996175e+12 1.7996196e+12  -35
1.7987937e+12 1.7987952e+12  -25
1863141360000 1863145440000  -68
1.5904071e+12 1.5904134e+12 -105
 1.608462e+12 1608469740000 -129
 1.615239e+12 1.6152498e+12 -180
1.6184745e+12 1.6184802e+12  -95
end
format %tc str_datetime
format %tc hosp_datetime
format %tc onset_hospitarrival

[William Lisowski]

Because the value produced by Clockdiff is a number of minutes, it does not need a datetime format. Also, you need to reverse the order of the first two arguments in your code.

. drop onset_hospitarrival

. generate onset_hospitarrival = Clockdiff(str_datetime, hosp_datetime , "minute")

. list, abbreviate(20) clean noobs

          str_datetime        hosp_datetime   onset_hospitarrival  
    09may2017 12:20:00   09may2017 14:35:00                   135  
    25jan2017 11:45:00   25jan2017 14:00:00                   135  
    23jan2017 18:00:00   23jan2017 18:20:00                    20  
    09jan2017 21:45:00   09jan2017 22:20:00                    35  
    31dec2016 08:55:00   31dec2016 09:20:00                    25  
    15jan2019 03:16:00   15jan2019 04:24:00                    68  
    25may2010 11:45:00   25may2010 13:30:00                   105  
    20dec2010 11:00:00   20dec2010 13:09:00                   129  
    08mar2011 21:30:00   09mar2011 00:30:00                   180  
    15apr2011 08:15:00   15apr2011 09:50:00                    95

[Van Ha]

Thanks! I used the format command because when I used the sum command on onset_hospitarrival, the output is in stata format that I can't make sense of... Does anyone know how to convert these values to ones that I can make sense of?

Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------------------
onset_hosp~l | 177,542 725.0271 49110.02 -535926 1.05e+07

[Clyde Schechter]

I take it you are not familiar with numbers in scientific notation, like 1.05e+07. If you are going to be doing statistical work on a regular basis, you should invest the time to learn about them. Read about the %e and %g formats in -help format- to do that.

However, if you -format onset_hospitarrival %10.0fc-, and then -summ onset_hospitarrival, format-, you will get numbers written out as integers with up to 8 digits, and commas separating groups of 3. The -format- option of -summarize- tells Stata to display the minimum and maximum in the same format that was applied to the variable itself.

All of that said, I wonder if there is a problem with your data in the first place. The maximum observed value, 1.05e+07 amounts to just a little under 20 years. Is that plausible? And the minimum value, -535926 is a little bit longer than a year. Is that plausible? And is it plausible for this variable to be negative at all? (Depends on what str_datetime represents, I suppose.)

[Van Ha]

Thanks Clyde for trying to help