corrected version of ts_hourly

[timmie]

small glitch in the function but mainly fixes in tests

[wesm]

BTW no need to create a new PR, you can just keep pushing commits to the branch you're working on

[changhiskhan]

@timmie, I implemented an hourly version in pivot_annual, pivot_annual and pivot_annual_h were so similar that it didn't make sense to have separate method and module just to have output in a different orientation. You can always use .T on the output.

I didn't merge in the extended_info and the default index. Here's why:

1. default index: having the wrong year is not the right way to do this. I guess you need it to access month/week/day properties. In that case you should make a function that can take a year and an array/index of ints that represents hours and create a correct datetimeindex.
2. extended info: would DataFrame.describe be enough?

I'm moving the rest of this issue to 0.11 if you still want to finish the 2 items I just listed above.

[timmie]

I implemented an hourly version in pivot_annual, pivot_annual and pivot_annual_h were so similar that it didn't

Thank you very much for that! Shall we add a similar part for minutely?

it didn't make sense to have separate method and module just to have output in a different orientation. You can always use .T on the output.

In earlier version I had problems with .T. And I flet that format easier to read.

default index: having the wrong year is not the right way to do this. I guess you need it to access month/week/day properties.

Exactly.

In that case you should make a function that can take a year and an array/index of ints that represents hours and create a correct datetimeindex.

Could you please show me an example? Python datetime always needs a year/month/day/hour.

extended info: would DataFrame.describe be enough?

well, I do not think so:

describe?
Generate various summary statistics of each column

So this is like the summary of a column.

extended info would generate the descriptives for all columns and add it to the dataframe

I.e. create long-term hourly averages.

Or would you do this also with DataFrame.describe?

Thanks again!

[wesm]

Is there anything remaining here or can be closed?

[changhiskhan]

@timmie, you can probably just do df.T.describe() to get most of what you want.

• minutely: sure, go ahead and just piggieback off of my code for that
• default index: maybe another way to do this is to generate the field info BEFORE you do the pivoting?
• datetime fields in extended info: I think maybe the right place for that is to make a method DatetimeIndex.fields that returns a DataFrame whose columns are the field names. This method should optionally take a list of field names.

I'm going to close this PR. Please make a new PR with the stuff mentioned above if you're interested.
As always, thanks for the contribution!