Import error

[anoukvlug]

It looks likes the pyclits package is precisely what I am looking for. However after an installation that appeared to be successful, I unfortunately I got the following error message once I tried to import the package:

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-1-b62475b6da31> in <module>
----> 1 import pyclits

~/Programs/miniconda3/envs/oggm_env/lib/python3.6/site-packages/pyclits/__init__.py in <module>
      5 """
      6 
----> 7 from data_loaders import *
      8 from empirical_model import EmpiricalModel
      9 from functions import *

ModuleNotFoundError: No module named 'data_loaders'

@jajcayn, could you help me solving this problem?

[jajcayn]

Dear @anoukvlug
unfortunately, pyclits is pretty old and unmaintained. The reason for your error is that pyclist was developed for python2.7 which is now deprecated. Last time I tried, it did not work with python3.

As for the future of pyclits, unfortunately I do not have the time right now to update it, but I started rewriting core functionality to python3 in branches rewrite/*. If you feel adventurous you might fork the package, checkout rewrite branch and try to finish basic core, then the installation on python3 should work.

Thanks for understanding,
Best,

Nikola

[anoukvlug]

Dear Nikola,
Thank you for your fast and kind reply! Unfortunately it is not the right time for me to be adventures and dive into the code.
Kind regards,
Anouk

[mafycek]

Dear Anouk, sidenote to the discussion. The new code that is in my branch = hynek_ refactoring_.... supports Python3. The Python environment is better contained using poetry that support simple packaging and publication of the library. Best, Hynek ---------- Původní e-mail ---------- Od: Anouk Vlug <notifications@github.com> Komu: jajcayn/pyclits <pyclits@noreply.github.com> Datum: 5. 6. 2020 12:53:18 Předmět: Re: [jajcayn/pyclits] Import error (#2) " Dear Nikola, Thank you for your fast and kind reply! Unfortunately it is not the right time for me to be adventures and dive into the code. Kind regards, Anouk — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub (#2 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACKGZPS3SCHL5O5SBA3IA5DRVDFBPANCNFSM4NTNXHXA) . "

[anoukvlug]

@mafycek thanks a lot for the advice! I just installed your branch and at first sight it seems to work :)

[mafycek]

Ok, cool! I did not know that it should work in my branch :) Btw. do you use virtual environment and poetry? ---------- Původní e-mail ---------- Od: Anouk Vlug <notifications@github.com> Komu: jajcayn/pyclits <pyclits@noreply.github.com> Datum: 5. 6. 2020 14:38:46 Předmět: Re: [jajcayn/pyclits] Import error (#2) " @mafycek(https://github.com/mafycek) thanks a lot for the advice! I just installed your branch and at first sight it seems to work :) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub (#2 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACKGZPQSOQBJCQCPKRS3Z7DRVDRNBANCNFSM4NTNXHXA) . "

[anoukvlug]

Sorry, I wrote my previous message to early. I thought you meant the branch would work with python 3. So after I installed both the branch and was able able to load the package, I thought it would work. However I kept running into errors, when trying to using the functions. It won't recognize all the functions (some it does recognize though) that are actually there:

e.g.:

AttributeError                            Traceback (most recent call last)
<ipython-input-19-f363db24ae29> in <module>
----> 1 nino34 = clt.data_loaders.load_enso_index("/aop1/avlug/nino34raw.txt", "3.4", start_date = date(1870, 1, 1), 
      2     end_date = date(2017, 7, 1), anom = False)
      3 

AttributeError: module 'pyclits' has no attribute 'data_loaders'

For this I used a python environment (env not venv). I'm not familiar with poetry.

[anoukvlug]

Actually, I wouldn't really need the data_loaders function, however I thought it would be good to get the example to work. As having an example would make it easier to figure out how to solve the error that I am having with the wavelet analysis function:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-23-4590a8ad37dc> in <module>
      1 wave, period, scale, coi = wavelet_analysis.continous_wavelet(df['001'].values, dt = dt, wavelet = mother, 
----> 2                                                               dj = dj, s0 = s0, j1 = j1, k=k0)

~/Programs/miniconda3/envs/oggm_env/lib/python3.6/site-packages/pyclits/wavelet_analysis.py in continous_wavelet(X, dt, pad, wavelet, **kwargs)
    137     k = np.arange(1, np.fix(n/2) + 1)
    138     k *= (2. * np.pi) / (n * dt)
--> 139     k_minus = -k[int(np.fix(n-1))/2 - 1::-1]
    140     k = np.concatenate((np.array([0.]), k, k_minus))
    141 

TypeError: slice indices must be integers or None or have an __index__ method

I have no clue if this error would be easy to solve or if it would get tricky and I can better look for another approach to do the wavelet analysis. @mafycek do you maybe know a quick fix for this problem? Or is this going to take quite some time to solve?

[mafycek]

I think that the problem is in the missing __init__ file. I never used the code but I could help you if you send me the command that teriggers the error. I need whole stack to investigate the issue.

…

---------- Původní e-mail ---------- Od: Anouk Vlug <notifications@github.com> Komu: jajcayn/pyclits <pyclits@noreply.github.com> Datum: 8. 6. 2020 12:40:42 Předmět: Re: [jajcayn/pyclits] Import error (#2) " Sorry, I wrote my previous message to early. I thought you meant the branch would work with python 3. So after I installed both the branch and was able able to load the package, I thought it would work. However I kept running into errors, when trying to using the functions. It won't recognize all the functions (some it does recognize though) that are actually there: e.g.: <code>AttributeError Traceback (most recent call last) <ipython-input-19-f363db24ae29> in <module> ----> 1 nino34 = clt.data_loaders.load_enso_index("/aop1/avlug/nino34raw.txt", "3.4", start_date = date(1870, 1, 1), 2 end_date = date(2017, 7, 1), anom = False) 3 AttributeError: module 'pyclits' has no attribute 'data_loaders' </code> For this I used a python environment (env not venv). I'm not familiar with poetry. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub (#2 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACKGZPQEBELPGJIHJ2F5K2TRVS5Z3ANCNFSM4NTNXHXA) . "

[mafycek]

Ok, here I see the problem. x/y return s float while x//y returns int. This is deviation beteen Python 3 and 2. ---------- Původní e-mail ---------- Od: Anouk Vlug <notifications@github.com> Komu: jajcayn/pyclits <pyclits@noreply.github.com> Datum: 8. 6. 2020 12:59:06 Předmět: Re: [jajcayn/pyclits] Import error (#2) " Actually, I wouldn't really need the data_loaders function, however I thought it would be good to get the example to work. As having an example would make it easier to figure out how to solve the error that I am having with the wavelet analysis function: <code>--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-23-4590a8ad37dc> in <module> 1 wave, period, scale, coi = wavelet_analysis.continous_wavelet(df['001'].values, dt = dt, wavelet = mother, ----> 2 dj = dj, s0 = s0, j1 = j1, k=k0) ~/Programs/miniconda3/envs/oggm_env/lib/python3.6/site-packages/pyclits/wavelet_analysis.py in continous_wavelet(X, dt, pad, wavelet, **kwargs) 137 k = np.arange(1, np.fix(n/2) + 1) 138 k *= (2. * np.pi) / (n * dt) --> 139 k_minus = -k[int(np.fix(n-1))/2 - 1::-1] 140 k = np.concatenate((np.array([0.]), k, k_minus)) 141 TypeError: slice indices must be integers or None or have an __index__ method </code> I have no clue if this error would be easy to solve or if it would get tricky and I can better look for another approach to do the wavelet analysis. @mafycek(https://github.com/mafycek) do you maybe know a quick fix for this problem? Or is this going to take quite some time to solve? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub (#2 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACKGZPQZE6NWYBQXYSHW5GDRVS77JANCNFSM4NTNXHXA) . "

[mafycek]

The problem should be fixed in the next commit.

…

---------- Původní e-mail ---------- Od: Anouk Vlug <notifications@github.com> Komu: jajcayn/pyclits <pyclits@noreply.github.com> Datum: 8. 6. 2020 12:59:06 Předmět: Re: [jajcayn/pyclits] Import error (#2) " Actually, I wouldn't really need the data_loaders function, however I thought it would be good to get the example to work. As having an example would make it easier to figure out how to solve the error that I am having with the wavelet analysis function: <code>--------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-23-4590a8ad37dc> in <module> 1 wave, period, scale, coi = wavelet_analysis.continous_wavelet(df['001'].values, dt = dt, wavelet = mother, ----> 2 dj = dj, s0 = s0, j1 = j1, k=k0) ~/Programs/miniconda3/envs/oggm_env/lib/python3.6/site-packages/pyclits/wavelet_analysis.py in continous_wavelet(X, dt, pad, wavelet, **kwargs) 137 k = np.arange(1, np.fix(n/2) + 1) 138 k *= (2. * np.pi) / (n * dt) --> 139 k_minus = -k[int(np.fix(n-1))/2 - 1::-1] 140 k = np.concatenate((np.array([0.]), k, k_minus)) 141 TypeError: slice indices must be integers or None or have an __index__ method </code> I have no clue if this error would be easy to solve or if it would get tricky and I can better look for another approach to do the wavelet analysis. @mafycek(https://github.com/mafycek) do you maybe know a quick fix for this problem? Or is this going to take quite some time to solve? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub (#2 (comment)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/ACKGZPQZE6NWYBQXYSHW5GDRVS77JANCNFSM4NTNXHXA) . "