'discharge' does not filter to only discharge about hydrofunctions

[mroberge]

This is a great idea! I love your solution. Would you be interested in submitting it in a pull request?

If you are interested, I could walk you through how to do it.

[mroberge]

I figured out what triggered this behavior. The default behavior is to return all of the data if no parameters are requested. So if you request my_nwis_object.df() you will get every parameter in the dataset. Because I set up separate pathways for 'discharge', 'q', and 'stage' (which I use all the time) the code is slightly different for these than for any other parameter. For these, if you request them and there is nothing there, no new columns get selected and df() goes to the next item (station.py line 238). For any other parameter, if the parameter isn't in one of the column names, it raises a ValueError (station.py line 251).

Some solutions are:

1. raise ValueErrors for 'discharge', 'q', & 'stage' at lines 238, 240, & 242 (con: it is repetitive, not DRY)
2. check list of args for 'discharge', 'q', & 'stage', then replace them with the parameter codes ( 00060 & 00065) (pro: it would be easier to add more shortcuts. con: adds another loop)
3. raise ValueError after the loop (line 272) instead of selecting all columns if no parameters are selected. (pro: if someone asks for two parameters and only one is there, it might be better to NOT raise an error? con: we'd have to find another way to return the whole dataframe if nothing is selected.)

[mroberge]

I think I like option # 2, which would just change the 'discharge' case from an elif to an if:

line 238:

    if item == 'discharge':
        item = '00060'
    elif item == 'stage':
        item = '00065'

... and so on. The other elif lines would stay the same, except for the first at line 248. This would substitute the shortcut for the actual parameter code, and then treat it like any other parameter code.

What are your thoughts?

BTW, would you like to add a 'res_level' shortcut or something?

[dgketchum]

I like solution #1. While it is repetitive, it will be easy to read/explicit. I'll set up a new PR. Thanks for your thoughts. Cheers,

…

On Tue, Jun 29, 2021 at 6:07 AM Martin Roberge ***@***.***> wrote: I figured out what triggered this behavior. The default behavior is to return all of the data if no parameters are requested. So if you request my_nwis_object.df() you will get every parameter in the dataset. Because I set up separate pathways for 'discharge', 'q', and 'stage' (which I use all the time) the code is slightly different for these than for any other parameter. For these, if you request them and there is nothing there, no new columns get selected and df() goes to the next item (station.py line 238). For any other parameter, if the parameter isn't in one of the column names, it raises a ValueError (station.py line 251). Some solutions are: 1. raise ValueErrors for 'discharge', 'q', & 'stage' at lines 238, 240, & 242 (*con*: it is repetitive, not DRY) 2. check list of args for 'discharge', 'q', & 'stage', then replace them with the parameter codes ( 00060 & 00065) (*pro*: it would be easier to add more shortcuts. *con*: adds another loop) 3. raise ValueError after the loop (line 272) instead of selecting all columns if no parameters are selected. (*pro*: if someone asks for two parameters and only one is there, it might be better to NOT raise an error? *con*: we'd have to find another way to return the whole dataframe if nothing is selected.) — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#102 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC4DUMV6GJQJ3WYNFODXCKTTVGZO5ANCNFSM47ORCBJA> .

-- David Ketchum Hydrologist, PhD Candidate Montana DNRC College of Forestry and Conservation, University of Montana Missoula, MT c. 8022227119

[mroberge]

@dgketchum I just took one of your commits from your pull request and added a test. I'm closing this issue with commit 86f83ec
Thank you for reporting this! I definitely have a blind spot- I mostly use discharge data, so I don't always notice problems involving other parameters. To help fix this, I added a new test fixture from your site in Lake Sherburne, Montana!