text
stringlengths 20
57.3k
| labels
class label 4
classes |
|---|---|
Title: Using type() to compare types - cannot use class that has `str` as base class
Body: It's not possible to use `StrEnum(str, Enum)` object as a key because of following code: https://github.com/aio-libs/aioredis/blob/231d0418988fd870c3d26d60eec65e2fc13d0144/aioredis/util.py#L35
Unfortunately this code requires argument to by strictly `str` type instead of being a subclass of `str`.
What I'd like to achieve is following:
```python
from enum import Enum
class StrEnum(str, Enum):
AAA = "aaa"
# ....
rdb.hmget("key", StrEnum.AAA)
```
| 0easy
|
Title: [ENH] implement LSTM model, and migrate `LSTMmodel` from notebook to main code base
Body: Just create this dataset:
```python
import numpy as np
import pandas as pd
multi_target_test_data = pd.DataFrame(
dict(
target1=np.random.rand(30),
target2=np.random.rand(30),
group=np.repeat(np.arange(3), 10),
time_idx=np.tile(np.arange(10), 3),
)
)
from pytorch_forecasting import TimeSeriesDataSet
from pytorch_forecasting.data.encoders import EncoderNormalizer, MultiNormalizer, TorchNormalizer
# create the dataset from the pandas dataframe
dataset = TimeSeriesDataSet(
multi_target_test_data,
group_ids=["group"],
target=["target1", "target2"], # USING two targets
time_idx="time_idx",
min_encoder_length=5,
max_encoder_length=5,
min_prediction_length=2,
max_prediction_length=2,
time_varying_unknown_reals=["target1", "target2"],
target_normalizer=MultiNormalizer(
[EncoderNormalizer(), TorchNormalizer()]
), # Use the NaNLabelEncoder to encode categorical target
)
```
And input it to the current `LSTMModel` in the tutorials:
```python
model = LSTMModel.from_dataset(
dataset,
n_layers=2,
hidden_size=10,
loss=MultiLoss([MAE() for _ in range(2)]),
)
x, y = next(iter(dataset.to_dataloader()))
print(
"prediction shape in training:", model(x)["prediction"].size()
) # batch_size x decoder time steps x 1 (1 for one target dimension)
model.eval() # set model into eval mode to use autoregressive prediction
print("prediction shape in inference:", model(x)["prediction"].size()) # should be the same as in training
```
And you'll get:
```
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
<ipython-input-13-037e20e20342> in <cell line: 3>()
2
3 print(
----> 4 "prediction shape in training:", model(x)["prediction"].size()
5 ) # batch_size x decoder time steps x 1 (1 for one target dimension)
6 model.eval() # set model into eval mode to use autoregressive prediction
/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py in _wrapped_call_impl(self, *args, **kwargs)
1530 return self._compiled_call_impl(*args, **kwargs) # type: ignore[misc]
1531 else:
-> 1532 return self._call_impl(*args, **kwargs)
1533
1534 def _call_impl(self, *args, **kwargs):
/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py in _call_impl(self, *args, **kwargs)
1539 or _global_backward_pre_hooks or _global_backward_hooks
1540 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1541 return forward_call(*args, **kwargs)
1542
1543 try:
<ipython-input-11-af0ffbd16c05> in forward(self, x)
105
106 def forward(self, x: Dict[str, torch.Tensor]) -> Dict[str, torch.Tensor]:
--> 107 hidden_state = self.encode(x) # encode to hidden state
108 output = self.decode(x, hidden_state) # decode leveraging hidden state
109
<ipython-input-11-af0ffbd16c05> in encode(self, x)
51 effective_encoder_lengths = x["encoder_lengths"] - 1
52 # run through LSTM network
---> 53 _, hidden_state = self.lstm(
54 input_vector, lengths=effective_encoder_lengths, enforce_sorted=False # passing the lengths directly
55 ) # second ouput is not needed (hidden state)
/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py in _wrapped_call_impl(self, *args, **kwargs)
1530 return self._compiled_call_impl(*args, **kwargs) # type: ignore[misc]
1531 else:
-> 1532 return self._call_impl(*args, **kwargs)
1533
1534 def _call_impl(self, *args, **kwargs):
/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py in _call_impl(self, *args, **kwargs)
1539 or _global_backward_pre_hooks or _global_backward_hooks
1540 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1541 return forward_call(*args, **kwargs)
1542
1543 try:
/usr/local/lib/python3.10/dist-packages/pytorch_forecasting/models/nn/rnn.py in forward(self, x, hx, lengths, enforce_sorted)
105 else:
106 pack_lengths = lengths.where(lengths > 0, torch.ones_like(lengths))
--> 107 packed_out, hidden_state = super().forward(
108 rnn.pack_padded_sequence(
109 x, pack_lengths.cpu(), enforce_sorted=enforce_sorted, batch_first=self.batch_first
/usr/local/lib/python3.10/dist-packages/torch/nn/modules/rnn.py in forward(self, input, hx)
912 self.dropout, self.training, self.bidirectional, self.batch_first)
913 else:
--> 914 result = _VF.lstm(input, batch_sizes, hx, self._flat_weights, self.bias,
915 self.num_layers, self.dropout, self.training, self.bidirectional)
916 output = result[0]
RuntimeError: mat1 and mat2 shapes cannot be multiplied (48x2 and 1x40)
```
It's just weird that there is still no fix for this, and no LSTM model out-of-the-box. I even made a fix, there is a [PR](https://github.com/jdb78/pytorch-forecasting/pull/1449).
Why does no one care about fixing this?
It is totally obscure how `pytorch_forecasting` handles uni-/multi-targets, I've also noticed that if you pass ` target=["target"]` to `TimeSeriesDataSet`, the `TimeSeriesDataSet` behaves very differently w.r.t. if you passed ` target="target"`.
Please just review that PR and even merge it, or fix it... | 0easy
|
Title: Remove things deprecated in 2.6.0
Body: https://docs.scrapy.org/en/latest/news.html#scrapy-2-6-0-2022-03-01 | 0easy
|
Title: [DOC] Docstring style should be made clear in `contributing` page
Body: # Brief Description of Fix
Currently, the docs do not clearly describe the coding standards that the project probably should adhere to.
I believe this should be added to the `Contributing` page.
For reference, it should look something like this:
<hr>
We recommend that you write docstrings before you write a test for the function, and that you write the test before writing the function implementation. This is an example of combining doc- and test-driven development together, to clarify what the function is intended to do before actually implementing it. Use the following example to help guide you on how to write your new function docstring properly.
```python
def new_janitor_function(df, param1, param2):
“””
One line that describes what the function does.
Further elaboration on the function. Use one or more paragraphs to do any of the following:
1. Provide more detail on intended usage
2. Clearly state assumptions on what the input dataframe should look like.
3. Identify edge cases that will cause the function to raise an error.
4. Other things that you want the end-user to know about before using the function.
Method chaining example:
.. code-block:: python
df = df.new_janitor_function(param1, param2)
Functional example:
.. code-block:: python
df = new_janitor_function(df, param1, param2)
:param param1: A description of param1.
:param param2: A description of param2.
:return: A description of the dataframe that is returned.
“””
# now put your function implementation here
```
# Relevant Context
- [Link to documentation page](https://pyjanitor.readthedocs.io/contributing.html#pull-request-guidelines)
- [Link to exact file to be edited](https://github.com/ericmjl/pyjanitor/blob/dev/CONTRIBUTING.rst)
| 0easy
|
Title: `Evaluate` keyword doesn't take attributes added into `builtins` module into account
Body: The keyword fails for a particular usecase as follows :
helper file:
```
import builtins
class dummy:
def __init__(self):
self.name = 'abc'
builtins.s = self
def foo(self):
print('in foo module')
def initialize(self):
print(self.name)
def dummy_initiliaze():
dummy().initialize()
```
Testcase file:
```
*** Settings ***
Library String
Library Collections
Library BuiltIn
Library helper.py
*** Test Cases ***
TC6
Dummy Initiliaze
${res} = Evaluate s.foo()
```
When I run this in Robot framework 3.1.2 , the function s.foo() is called and works
But in Robot framework 3.2.2, its erroring out with "Evaluating expression 's.foo()' failed: NameError: name 's' is not defined nor importable as module"
Have discussed the same in slack forum, thread : https://robotframework.slack.com/archives/C3C28F9DF/p1718686445500879
| 0easy
|
Title: RoBERTa on SuperGLUE's BoolQ task
Body: BoolQ is one of the tasks of the [SuperGLUE](https://super.gluebenchmark.com) benchmark. The task is to re-trace the steps of Facebook's RoBERTa paper (https://arxiv.org/pdf/1907.11692.pdf) and build an AllenNLP config that reads the BoolQ data and fine-tunes a model on it. We expect scores in the range of their entry on the [SuperGLUE leaderboard](https://super.gluebenchmark.com/leaderboard).
We recommend you use the [AllenNLP Repository Template](https://github.com/allenai/allennlp-template-config-files) as a starting point. It might also be helpful to look at how the TransformerQA model works ([training config](https://github.com/allenai/allennlp-models/blob/main/training_config/rc/transformer_qa.jsonnet), [reader](https://github.com/allenai/allennlp-models/blob/main/allennlp_models/rc/dataset_readers/transformer_squad.py), [model](https://github.com/allenai/allennlp-models/blob/main/allennlp_models/rc/models/transformer_qa.py)). | 0easy
|
Title: Apply Black to docs
Body: It may be a matter of adding https://github.com/adamchainz/blacken-docs to out pre-commit config. | 0easy
|
Title: Error: Algorithm should either be a classifier to be used with response_method=predict_proba or the response_method should be 'predict'
Body: Hi!
Recently I started getting this error (or warning?) with binary classification.
Minimal code to reproduce the problem:
```
from supervised import AutoML
if __name__ == '__main__':
from sklearn.datasets import make_classification
X, y = make_classification(n_samples=100000, n_features=20, n_redundant=2)
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.3)
automl = AutoML(eval_metric="accuracy")
automl.fit(X_train, y_train)
automl.report()
```
I get multiple messages:
```
DecisionTreeAlgorithm should either be a classifier to be used with response_method=predict_proba or the response_method should be 'predict'. Got a regressor with response_method=predict_proba instead.
Problem during computing permutation importance. Skipping ...
```
Is it a new bug or we can just ignore it?
Thank you!
| 0easy
|
Title: Misdetected data type when numerical column contains null
Body: When a column contains null values, it can get misclassified as a nominal type, even if it has high cardinality and is largely quantitative. For instance, in this example, the `# Instances` and `# Attributes` are detected as `nominal` but they are better suited as `quantitative`.
```python
import pandas as pd
import lux
df = pd.read_html("https://archive.ics.uci.edu/ml/datasets.php?format=&task=&att=&area=&numAtt=&numIns=&type=&sort=nameUp&view=table")[5]
df.columns = df.loc[0]
df = df.loc[1:]
df['Year'] = pd.to_datetime(df['Year'], format='%Y')
df.data_type
```
As a result, even when dropna occurs on these columns, the data type remains `nominal`.
```python
df[['# Instances','# Attributes']].dropna()
```
| 0easy
|
Title: Removed notebooks remain in the Mercury
Body: I'm running Mercury locally. I had `eda.ipynb` file in the directory. It was served fine in the Mercury. I renamed the notebook file. Now, in the Mercury I had two notebooks, with old and new name. The notebook with old name was not working in the Mercury. The new one is working fine.
There should be some check if notebook is still available when starting Mercury locally. | 0easy
|
Title: A new test function from Oakley et al. (2004)
Body: In the following paper, Oakley et al. (2004) propose a test function of the form
The data for the test function is available on the author's [website](http://www.jeremy-oakley.staff.shef.ac.uk/psa_example.txt). The analytical results are in the paper (table 1 on page 764).
Oakley, Jeremy E., and Anthony O’Hagan. “Probabilistic Sensitivity Analysis of Complex Models: A Bayesian Approach.” Journal of the Royal Statistical Society: Series B (Statistical Methodology) 66, no. 3 (August 2004): 751–769. doi:10.1111/j.1467-9868.2004.05304.x.
| 0easy
|
Title: Move bandit to dev section
Body: It seems `bandit` is only used in the development stage.
Does it make sense to move the dependency to the `dev` section on [pyproject.toml](https://github.com/scanapi/scanapi/blob/master/pyproject.toml#L26)? | 0easy
|
Title: 'json' selector type not documented
Body: https://docs.scrapy.org/en/latest/topics/selectors.html
`type defines the selector type, it can be "html", "xml" or None (default).`
But my tests are now failing because it's returning `json` type for actual json docs now. | 0easy
|
Title: [FEATURE] Run with given regressor instead of raising warning in ZeroInflatedRegressor
Body: I am using the ZeroInflatedRegressor for the prediction of some materials' prices. For some materials I am getting this error:
```
Traceback (most recent call last):
File "/home/bugra/project-mats/mat_analyser.py", line 61, in apply_regression
zir.fit(X_train, y_train)
File "/home/bugra/project-mats/mats_folder/lib/python3.7/site-packages/sklego/meta/zero_inflated_regressor.py", line 107, in fit
"The predicted training labels are all zero, making the regressor obsolete. Change the classifier or use a plain regressor instead.")
ValueError: The predicted training labels are all zero, making the regressor obsolete. Change the classifier or use a plain regressor instead.
```
Tbh. I didn't understand much from the error. But, I guess using the regressor (in my case this is LinearRegression from sklearn) could give me some results. Why ZeroInflatedRegressor can't use the regressor I defined for it but instead raising this error? I think there should be a flag (a parameter) in ZeroInflatedRegressor which would let it fit the regressor as suggested by the error message if the user wants. If not, it could still raise this error.
| 0easy
|
Title: FEAT: support `Rolling.cov`
Body: ### Is your feature request related to a problem? Please describe
`Rolling.cov` calculates the rolling sample covariance. It is a common operator in quant trading workflow.
### Describe the solution you'd like
`Rolling.cov` could be implemented in a way similar to `Rolling.corr`
| 0easy
|
Title: [4.12.0][regression] The presence of `TOX_PARALLEL_NO_SPINNER=1` in CI suppresses output of single env run tox commands
Body: ## Issue
I have a number for globally set environment variables, one of each is `TOX_PARALLEL_NO_SPINNER`: https://github.com/ansible/pylibssh/blob/ef77a85/.github/workflows/ci-cd.yml#L75.
In one of the steps, I run `tox -e a-single-env -qq | tee a-file.txt`. In tox 4.11.4, it produces output from the command specified in `tox.ini`. Starting with tox 4.12.0, the output is empty.
I'm pretty convinced this is a side effect of #3159.
Previously, setting this var allowed having it in one place, while allowing for parallel and non-parallel runs different job steps, not having to track whether to disable or enable the spinner in each. So I consider this to be a regression.
## Environment
- OS: ubuntu-latest @ GHA
`tox == 4.12.0`
## Output of running tox
No output.
## Minimal example
```console
$ TOX_PARALLEL_NO_SPINNER=1 tox -e a-single-env --skip-pkg-install -qq
```
| 0easy
|
Title: Non-deterministic Import Ordering
Body: **Describe the bug**
The **update** command when executed with an unchanged openapi yaml file multiple times results in template model and operation files with different orders of **imports** at the top of the file. If the resulting client is stored in a git repository, this causes unecessary differences between commits. _"Non-deterministic"_ is intended in the meaning of the **order** of the imports rather than the **set** which appears to be the same regardless of the number of update commands.
**To Reproduce**
Steps to reproduce the behavior:
1. Generate a client from an openapi yaml file.
2. Commit the output client to a git repo (optional, but makes the symptom easier to see)
3. Update a client from the same, unchanged openapi yaml file.
4. Run a git diff to see the changed import ordering (usually in almost every generated file)
**Expected behavior**
An order-preserved set of import statements for each generated file.
**OpenAPI Spec File**
Any OpenAPI yaml file.
**Desktop (please complete the following information):**
- OS: Ubuntu Desktop 20
- Python Version: 3.8.x
- openapi-python-client version: 0.10.1 ( using pipx run openapi-python-client )
| 0easy
|
Title: Feature Proposal: Auto Agg method
Body: # Brief Description
An `auto_agg` method where you just put the columns you want and it automatically aggregates the dataframe.
For example, I could do something like:
```python
df.auto_agg(['state', 'age_mean'])
```
On a dataframe that has a `state` and `age` column and it will automatically get the average age by state.
# Example API
Here's some basic code:
```python
import pandas as pd
colnames = ('id', 'state', 'state_abbr', 'amount')
df = pd.DataFrame.from_records((
('Person1', 'California', 'CA', 20),
('Person2', 'California', 'CA', 30),
('Person3', 'New York', 'NY', 15),
('Person4', 'New York', 'NY', 10),
('Person5', 'New York', 'NY', 45)), columns=colnames)
new_columns = ['state_abbr', 'state_count']
def flatten_dict(x):
return {k: v for d in x for k, v in d.items()}
def auto_agg(df, new_columns):
cols = set(df.columns)
new_columns_set = set(new_columns)
group_cols = cols.intersection(new_columns_set)
agg_cols = new_columns_set.difference(cols)
agg_groups = [{col: agg_col} for col in cols for agg_col in agg_cols if agg_col.startswith(col)]
this_dict = dict()
for i in agg_groups:
value = list(i.values())[0]
key = list(i.keys())[0]
this_dict[value] = ''
current_value = this_dict.get(value)
if len(key) > len(current_value):
this_dict[value] = key
agg_groups_rename = dict([(value, key) for key, value in this_dict.items()])
agg_groups_agg = [{i[0]: i[1].split(i[0])[1:][0][1:]} for i in agg_groups_rename.items()]
agg_groups_agg = flatten_dict(agg_groups_agg)
group_cols_list = list(group_cols)
return df.groupby(group_cols_list, as_index=False).agg(agg_groups_agg).rename(columns=agg_groups_rename)
auto_agg(df, new_columns)
```
There's a problem when column names are subsets of eachother, which is why I have an example with `state_abbr` and `state`. Sometimes the set math gets confused when trying to split the column names and the functions.
I think a basic workaround would be to be able to "force" the functions on specific columns without us needing to guess, like this:
```python
new_columns = ['state', 'amount_sum', {'id': 'count'}]
auto_agg(df, new_columns)
# or df.auto_agg(new_columns)
```
If we say something like "if your choice of columns mess up, then be explicit" we might not need this part of the code, since we won't need to find the biggest or smallest version of the column name.
```python
for i in agg_groups:
value = list(i.values())[0]
key = list(i.keys())[0]
this_dict[value] = ''
current_value = this_dict.get(value)
if len(key) > len(current_value):
this_dict[value] = key
``` | 0easy
|
Title: [UX][API Server] Spinner not shown on the stream webpage
Body: The spinner seems not rendered in the api streaming. Examples:
<img width="992" alt="Image" src="https://github.com/user-attachments/assets/daa7a5ec-3cb2-46f7-bc08-ff481e580e1b" />
<img width="989" alt="Image" src="https://github.com/user-attachments/assets/33e7fede-5ef2-4206-a420-4e5ae472d5b8" /> | 0easy
|
Title: [BUG] No activations in BlockRNNModel output MLP
Body: **Describe the bug**
I've been trying to understand the differences between `BlockRNNModel` and `RNNModel`; in particular, why one only supports future covariates, one only supports past covariates, and neither support static covariates. This is still unclear to me, however it's not the reason for this issue.
While I was looking at the source, I noticed that in `BlockRNNModel`, an MLP is being created which is just a stack of linear layers with no activations in between, which seems like a mistake.
https://github.com/unit8co/darts/blob/a646adf10fa73d8facb16a611f8a3682dc8a1191/darts/models/forecasting/block_rnn_model.py#L158-L163
**Expected behavior**
There should be an nonlinear activation function between the linear layers, otherwise it is the same as just having a single linear layer. An argument to `BlockRNNModel` should be provided that allows the user to choose this activation function. | 0easy
|
Title: Replace _compat.cached_property with functools.cached_property
Body: **Description of the issue**
cirq uses an internal version of cached_property defined here to support earlier versions of Python.
https://github.com/quantumlib/Cirq/blob/d33b1a71ac9721e762a2fa1b9e9871edfe187a3c/cirq-core/cirq/_compat.py#L64-L67
As of #6167 we require Python 3.9 which is guaranteed to have [functools.cached_property](https://docs.python.org/3/library/functools.html#functools.cached_property).
We can therefore replace all instances of `_compat.cached_property` with `functools.cached_property` and remove the `cached_property` definition from the _compat.py.
**Cirq version**
1.4.0 at d33b1a71ac9721e762a2fa1b9e9871edfe187a3c | 0easy
|
Title: Advanced Encodings for Categoricals
Body: Add Ability to Category Encode : include encoding with KFold too. | 0easy
|
Title: Effect Plot for Linear Models
Body: **Describe the solution you'd like**
Would love to have an Effect Plot for aiding with interpreting linear models. I realize that a feature importance plot does some of this. An effect plot shows the weights as a bar plot so you can see whether the impact is positive or negative and also how large the variance is.
**Examples**
There is a great example here
https://christophm.github.io/interpretable-ml-book/limo.html#visual-parameter-interpretation
My scouring has not turned up any Python code to generate this plot in the wild.
| 0easy
|
Title: Still getting "cannot import name 'NaN' from numpy" error
Body: **Which version are you running? The lastest version is on Github. Pip is for major releases.**
```python
import pandas_ta as ta
print(ta.version)
```
Version: 0.3.14b0
**Do you have _TA Lib_ also installed in your environment?**
```sh
$ pip list
```
no. should i?
**Have you tried the _development_ version? Did it resolve the issue?**
```sh
$ pip install -U git+https://github.com/twopirllc/pandas-ta.git@development
```
yes and it did not solve it.
**Describe the bug**
A clear and concise description of what the bug is.
Still getting "cannot import name 'NaN' from numpy" error and have tried suggested solutions with no luck.
**To Reproduce**
Provide sample code.
```python
import pandas as pd
import pandas_ta as ta
file_path = r'C:\Users\micha\OneDrive\Desktop\Algo Switcher Model\combined_output.txt' # Replace with the actual path to your combined_output.txt
data = pd.read_csv(file_path, delimiter=';', header=None)
data['ATR'] = ta.atr(data['High'], data['Low'], data['Close'], length=14)
data['ADX'] = ta.adx(data['High'], data['Low'], data['Close'], length=14)['ADX_14']
data['+DI'] = ta.adx(data['High'], data['Low'], data['Close'], length=14)['DMP_14']
data['-DI'] = ta.adx(data['High'], data['Low'], data['Close'], length=14)['DMN_14']`
```
**Expected behavior**
A clear and concise description of what you expected to happen.
import pandas_ta should work !
**Screenshots**
If applicable, add screenshots to help explain your problem.
**Additional context**
Add any other context about the problem here.
Changed the squeeze pro file to "from numpy import nan" from "from numpy import NaN as npNaN"
also tried having it as "from numpy import nan as npNaN"
neither solution worked.
Running in visual studio 2022. Python 3.9 64-bit.
| 0easy
|
Title: Feature: allow including a`faststream.broker.router.BrokerRouter()` in a `StreamRouter`
Body: **Is your feature request related to a problem? Please describe.**
Consider the following scenario: You have an existing module that's based on a BrokerRouter (e.g. a NatsRouter) and it only has subscribers and publishers. You want to be able to re-use that module in plain FastStream apps and other apps (e.g. a FastAPI app). This means that at the top-level, you need to do something like
```
app_router = faststream.nats.fastapi.NatsRouter() # Any subclass of StreamRouter
app_router.include_router(shared_broker_routes.router) # This raises because `BrokerRouter.routes` doesn't exist.
app_router.include_router(shared_api_routes.router)
app = FastAPI()
# ...
```
**Describe the solution you'd like**
It would be nice to not have to change existing BrokerRouter routes
a few options might be
- change `StreamRouter.include_router` to conditionally call `super().include_router()`
- change BrokerRouter to expose an empty routes property (this would work for Nats + FastAPI / Starlette but I'm not sure how universal that would be.
- Attempt to unify the concepts of BrokerRouter handlers and and APIRouter routes
**Describe alternatives you've considered**
I currently can't directly share the code, I have to force everything to be a fastapi StreamRouter.
| 0easy
|
Title: Missing webconfig/js/xonsh_sticker.svg in xonsh package
Body: `xonfig web` shows [empty image](https://github.com/xonsh/xonsh/blob/204bb9907962276a476b2f07e1dc153d3c37093c/xonsh/webconfig/index.html#L14):
<img width="230" alt="image" src="https://user-images.githubusercontent.com/1708680/219278218-08f245e8-153a-4365-b0fd-dbfd4763fa07.png">
The image [xonsh_sticker.svg](https://github.com/xonsh/xonsh/blob/main/xonsh/webconfig/js/xonsh_sticker.svg) exists in repo.
But if you [download the package](https://pypi.org/project/xonsh/#files) there is no [xonsh_sticker.svg](https://github.com/xonsh/xonsh/blob/main/xonsh/webconfig/js/xonsh_sticker.svg) in `webconfig/js/`.
## For community
⬇️ **Please click the 👍 reaction instead of leaving a `+1` or 👍 comment**
| 0easy
|
Title: Original order of dictionaries is not preserved when they are pretty printed in log messages
Body: Python 3.9.2
robotframework 6.0.2
Use case example:
I would like to keep the order of &{targets} from keyword call. For now Free Named Arguments are sorted by alphabetical order.
| 0easy
|
Title: Verify notebook-based pip install
Body: I'm not familiar with the `collab` notebook environment. I'm looking at all of our notebooks, and I see that we have text like `pip install hummingbird_ml[extra]` at the top of all of them. This was added by a `collab` user (ex: #150).
I would expect this to be `pip install hummingbird-ml[extra]` with a dash, not underscore...but before I change it back maybe there is a notebook environment difference that I am not aware of!
Could someone who is familiar with `collab` test our pip install import on both collab and jupyter in our notebook examples? And change it if necessary or close this issue if not?
Thanks!! | 0easy
|
Title: No way to copy selected text ?
Body: Hello,
Kudos for the great job !
I've got some issues though :
Using pywinauto I navigate to an input text.
Using type_keys (^a^c) I can select the text, and copy it to the clipboard (I can paste it using ctrl+v manually).
But there doesn't seem to be a way to retrieve it programmatically.
Or am I mistaken ?
In the doc there is a mention to pywinauto.clipboard, but it doesn't work at all for me :
```
>python -m pip install pywinauto --upgrade
Requirement already up-to-date: pywinauto in python\python36\lib\site-packages (0.6.4)
...
>python
Python 3.6.0a1 (v3.6.0a1:5896da372fb0, May 17 2016, 16:21:39) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import pywinauto
>>> pywinauto.clipboard
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: module 'pywinauto' has no attribute 'clipboard'
```
Any help would be much appreciated, thank you for your time. | 0easy
|
Title: Upstream Code Dependencies metric API
Body: The canonical definition is here: https://chaoss.community/?p=3977 | 0easy
|
Title: [FEA] Line graph transform and enrichments
Body: **Is your feature request related to a problem? Please describe.**
Graph compute libraries like `cuGraph` support all sorts of node-centric community detectors, but generally not edge-centric ones, which are needed for SNA. The line graph transform, esp. GPU-accelerated, would simplify reusing node-centric classifiers for edges, especially if paired with reverse projection method.
**Describe the solution you'd like**
1. Methods `to_line_graph`, `from_line_graph`, and maybe `enrich_from_g`
```python
g = graphistry.edges(...)
lg = g.to_line_graph()
lg2 = ... cuGraph.communtity ...
g2 = lg2.from_line_graph()
g2b = g.enrich_from_g(node_attribs=[..], edge_attribs=[...])
```
2. Docs example featuring community detection on edges:
- transform
- edge community
- reverse transform
- propagation to node multi-label community classification
**Describe alternatives you've considered**
- Explicit line graph type: doesn't seem necessary
- more generic `from_line_graph`: doesn't seem necessary
**Additional context**
The prime motivator is really for community detection algs, but worth checking if others
| 0easy
|
Title: analyze methods do not allow Y as a list
Body: currently required to be a numpy array.
Could add a simple check at the beginning of every method, `if type(Y) is list: Y = np.array(Y)`
| 0easy
|
Title: Marketplace - creator page - change the padding on the top categories chips
Body:
### Describe your issue.
Can we adjust the padding on these chips?
Please change the padding to:
top and bottom padding to 12px
left and right padding to: 16px
<img width="526" alt="Screenshot 2024-12-16 at 20 40 59" src="https://github.com/user-attachments/assets/47287a50-5065-48e3-a8e3-9235343b1bb3" />
### Upload Activity Log Content
_No response_
### Upload Error Log Content
_No response_ | 0easy
|
Title: Docs: Add common recipes/FAQ section
Body: ### Reported by
[Coffee](https://discord.com/users/489485360645275650) in Discord: Unable to see stack trace inside my routes
### Description
we need a sticky, or FAQ or something
### MCVE
User asked in discord how to enable visibility into stacktraces, is a frequently asked question.
### Logs
N/A
### Litestar Version
All | 0easy
|
Title: Documentation needs to be updated for LM Studio
Body: ### Describe the bug
There are two sections of the documentation for running local models and they both should be updated
The first is under `Guides` on a page called `Running Locally`. This currently only includes instructions for using open source services. I suggest we include instructions for running LM Studio even though it isn't in the included video, because of how popular LM Studio is.
The second is under `Language Models` in a section called `Local Providers`. All providers have their own dedicated page, including LM studio, which could use the following improvements:
- [ ] Make text/copy specific to LM Studio. Some text seems to target all local providers instead of LM Studio specifically.
- [ ] Document when users include `model` flag
- [ ] No longer the default when running `interpreter --local`
- [ ] Probably more stuff. I encourage you to play with OI using LM Studio and try different commands/flags to see what happens and include it in the docs
### Reproduce
Go to
- https://docs.openinterpreter.com/language-models/local-models/lm-studio
- https://docs.openinterpreter.com/guides/running-locally
Observe the docs
### Expected behavior
Docs accurately reflect the functionality of the latest version of OI
### Screenshots
_No response_
### Open Interpreter version
0.2.2
### Python version
3.11.3
### Operating System name and version
All
### Additional context
_No response_ | 0easy
|
Title: CSS/JS console and ipython fix
Body: Some of our code snippets display as a terminal (i.e., the line starts with a `$`), or an IPython prompt (i.e., line starts with `In []:`). See this: https://ploomber.readthedocs.io/en/latest/user-guide/scaffold.html
The corresponding CSS is here: https://github.com/ploomber/ploomber/blob/fd9b4c7a2e787c0206f841928d1be90ac142c7a8/doc/_static/css/custom-theme.css#L83
However, this only works when the terminal only displays one line, if there are more lines, it doesn't work:
```sh
$ some command
another_command # this one doesn't have the $
```
Not sure if we can fix this purely with CSS or we need some JS.
| 0easy
|
Title: [ENH] Add Spline Forecaster
Body: **Is your feature request related to a problem? Please describe.**
So far, when detrending, we can use linear regression with polynomial features. I would like something more flexible, i.e. fiting linear splines in a way that Prophet does it, but without the probabilistic programming overhead and with more control about what happens when extrapolating.
**Describe the solution you'd like**
I would like a new class like the `PolynomialTrendForecaster`, but `SplineTrendForecaster` or so. Internally, it would look like the `PolynomialTrendForecaster`, with (hopefully) the only difference being that instead of
```python
self.regressor_ = make_pipeline(
PolynomialFeatures(degree=self.degree, include_bias=self.with_intercept),
regressor,
)
```
we would have something like
```python
self.regressor_ = make_pipeline(
SplineTransformer(degree=self.degree, n_knots=self.knots, extrapolation=self.extrapolation, include_bias=self.with_intercept),
regressor,
)
```
We could maybe even write another class `PipelineTrendForecaster` that takes a scikit-learn pipeline. From that, we could easily derive `PolynomialTrendForecaster` and `SplineTrendForecaster`.
scikit-learn's `SplineTransformer` has the advantage that you can tell it how to extrapolate, either linearly, constant, periodic, or more. | 0easy
|
Title: Add Cohere docs
Body: Looks like `Cohere` is missing install + other docs. | 0easy
|
Title: onnx-weekly has inconsistent version metadata
Body: # Bug Report
### Is the issue related to model conversion?
<!-- If the ONNX checker reports issues with this model then this is most probably related to the converter used to convert the original framework model to ONNX. Please create this bug in the appropriate converter's GitHub repo (pytorch, tensorflow-onnx, sklearn-onnx, keras-onnx, onnxmltools) to get the best help. -->
No
### Describe the bug
<!-- Please describe the bug clearly and concisely -->
Compiling the onnx-weekly package during installation fails due to inconsistent meta data:
```
$ pip install --no-binary :all: onnx-weekly
Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com
Collecting onnx-weekly
Downloading onnx-weekly-1.16.0.dev20240205.tar.gz (12.3 MB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 12.3/12.3 MB 8.2 MB/s eta 0:00:00
Installing build dependencies ... done
Getting requirements to build wheel ... done
Installing backend dependencies ... done
Preparing metadata (pyproject.toml) ... done
Discarding https://files.pythonhosted.org/packages/20/f2/bf15d54fd4e661f1e5f3df79fbbbbc32ac5e4912154ed0ea6f2854dc3540/onnx-weekly-1.16.0.dev20240205.tar.gz (from https://pypi.org/simple/onnx-weekly/) (requires-python:>=3.8): Requested onnx-weekly from https://files.pythonhosted.org/packages/20/f2/bf15d54fd4e661f1e5f3df79fbbbbc32ac5e4912154ed0ea6f2854dc3540/onnx-weekly-1.16.0.dev20240205.tar.gz has inconsistent version: expected '1.16.0.dev20240205', but metadata has '1.16.0'
```
Building the source package directly from the tarball works though:
`pip install https://files.pythonhosted.org/packages/20/f2/bf15d54fd4e661f1e5f3df79fbbbbc32ac5e4912154ed0ea6f2854dc3540/onnx-weekly-1.16.0.dev20240205.tar.gz`
What also works is installation with the older resolver engine:
`pip install --use-deprecated=legacy-resolver --no-binary :all: onnx-weekly`
### System information
<!--
- OS Platform and Distribution (*e.g. Linux Ubuntu 20.04*):
- ONNX version (*e.g. 1.13*):
- Python version:
- GCC/Compiler version (if compiling from source):
- CMake version:
- Protobuf version:
- Visual Studio version (if applicable):-->
- Fedora 38 on x86-64
### Reproduction instructions
<!--
- Describe the code to reproduce the behavior.
```
import onnx
model = onnx.load('model.onnx')
...
```
- Attach the ONNX model to the issue (where applicable)-->
`pip install --no-binary :all: onnx-weekly`
### Expected behavior
<!-- A clear and concise description of what you expected to happen. -->
Build and install onnx-weekly package
### Notes
<!-- Any additional information -->
| 0easy
|
Title: Add SEO fields to product and category models and make editable in the dashboard
Body: There are a few SEO components that we regularly find clients asking for the ability to control on product and category pages:
1. Meta title
2. Meta description
3. Slug
For products, slug, meta title and description are all auto-generated from the product title/description, with no option to override any of these. While this is fine a lot of the time, Oscar should provide the ability to override these.
For categories, meta title and description are also auto-generated from the product title/description, and there is no option to override.
I think we should have a "Search Engine Optimisation" tab on the dashboard edit views for both of these, which allow override of these parameters. | 0easy
|
Title: JMA indicator ValueError: cannot convert float NaN to integer
Body: pandas-ta: 0.3.14b0
Running `df.ta.strategy()` or more specifically `df.ta.jma()` on a simple dataframe fails with
```sh
Error
Traceback (most recent call last):
File "/Users/andrei/Projects/BE/breakingequity/breakingequity-backtest-single-day/tests/test_ohlcdata.py", line 60, in test_jma
df.ta.jma()
File "/Users/andrei/.local/share/virtualenvs/breakingequity-backtest-single-day-Vn0CDZ8S/lib/python3.9/site-packages/pandas_ta/core.py", line 1199, in jma
result = jma(close=close, length=length, phase=phase, offset=offset, **kwargs)
File "/Users/andrei/.local/share/virtualenvs/breakingequity-backtest-single-day-Vn0CDZ8S/lib/python3.9/site-packages/pandas_ta/overlap/jma.py", line 76, in jma
jma[0:_length - 1] = npNaN
ValueError: cannot convert float NaN to integer
```
Sample unit test:
```python
def test_jma(self):
minutes_range = pd.date_range(
start=pd.Timestamp('2020-01-02 09:30:00-05:00'),
end=pd.Timestamp('2020-01-02 16:00:00-05:00'),
freq='1min')
minutes = 390
ohlc = [
[1, 2, 1, 2, 10, 1.5],
[2, 4, 2, 4, 20, 3.0],
[3, 4, 3, 4, 20, 3.0],
]
df = pd.DataFrame(
ohlc * int(minutes / len(ohlc)),
columns=['open', 'high', 'low', 'close', 'volume', 'vwap'],
index=minutes_range[0:minutes]
)
df.ta.jma()
```
| 0easy
|
Title: Unused assigned variable (remove this)
Body: https://github.com/mithi/hexapod-robot-simulator/blob/030cdf7b6b293fc38a04ca8a68ee2527f1f4c8d7/hexapod/ik_solver/recompute_hexapod.py#L43 | 0easy
|
Title: All numeric registration code does not work
Body: Registration codes that are all numeric don't work b/c we automatically cast it to an int.
This code needs a call to str so that we get a string always.
https://github.com/CTFd/CTFd/blob/514ab2c8bd3b0687615307c19fa9618b09e3998a/CTFd/auth.py#L214-L219 | 0easy
|
Title: Test that transform_column does not mutate original
Body: Linked from issue #645.
> Should be put a test somewhere for some garantee?
>
> _Originally posted by @VPerrollaz in https://github.com/ericmjl/pyjanitor/issues/645#issuecomment-630170156_
Possible test outline (just pseudo code, don't treat it as real code):
```python
def test_transform_column_does_not_mutate_original():
df_transformed = df.transform_column("some_col", some_func)
with pytest.raises(AssertionError):
pd.testing.assert_frame_equal(df, df_transformed)
``` | 0easy
|
Title: [docs] Improve nginx example configuration in the deploy FAQ
Body: The example [nginx configuration](https://falcon.readthedocs.io/en/latest/deploy/nginx-uwsgi.html#id3 ) in the `Deploying Falcon on Linux with NGINX and uWSGI` article could probably be improved to include https configuration and redirect of http to https, since I believe it's the most common configuration now.
Maybe we could use a configuration from this great [Mozilla site](https://ssl-config.mozilla.org/#server=nginx) and link to it since it also has ssl configuration for many other common programs? | 0easy
|
Title: [Proposal] Add Tutorials for MuJoCo based environments (contributors welcome)
Body: ### Proposal
With the release of the `MuJoCo-v5` environments, in Gymnasium 1.0.0 (which will be coming out, prior to the heat death of the universe).
We need tutorials on:
- [x] loading a quadruped custom robot model (WIP https://github.com/Farama-Foundation/Gymnasium/pull/838)
- [ ] loading a bipedal / humanoid custom robot model
- [ ] loading a custom scene e.g. [G-Barkour](https://github.com/google-deepmind/mujoco_menagerie/blob/main/google_barkour_v0/README.md) (which will have custom reward, obs?, terminal?)
- [ ] Using robot model with custom simulation options
- [ ] Create a Custom MuJoCo based environment class by subclassing `MujocoEnv`
- [ ] ??? Transfer learning across different models ???
- [ ] Transfer learning with [MuJoCo MPC](https://github.com/google-deepmind/mujoco_mpc/)
- [ ] more tutorials ???
Contributors are welcome, with experience running RL on robotic environments, or with knowledge of the `MuJoCo simulator`, or with knowledge of Robotics, or just willingness to learn.
Comment here or discuss or join the gymnasium discord (invite link on the README) and ping me in the `# New Contributors` channel.
### Motivation
_No response_
### Pitch
_No response_
### Alternatives
_No response_
### Additional context
_No response_
### Checklist
- [X] I have checked that there is no similar [issue](https://github.com/Farama-Foundation/Gymnasium/issues) in the repo
| 0easy
|
Title: Enforce "decode_responses=True" when connecting to Redis?
Body: **Is your feature request related to a problem? Please describe.**
After scratching my head for an hour or more why a unit test of mine fails (but everything is fine in production) I finally noticed that is configures Dynaconf like this:
```
settings.configure(
ENV_FOR_DYNACONF="testing",
ROOT_PATH_FOR_DYNACONF=os.path.dirname(os.path.realpath(__file__)) + "/../../../config",
REDIS_ENABLED_FOR_DYNACONF=True,
# Actual values are not important - fakeredis is around
REDIS_FOR_DYNACONF={"host": "irrelevant", "port": 1111},
)
```
While this does not look too bad at first sight it silently runs into a problem: The (useless) `REDIS_FOR_DYNACONF` parameter overrides the Dynaconf internal default which contains `"decode_responses": get("REDIS_DECODE_FOR_DYNACONF", True)`.
This means all keys received from Redis are now bytes and not the expected strings.
As `dynaconf.loaders.redis_loader.load` is called with `silent=True` by default no data is read from Redis as the following exception is swallowed:
```
E TypeError: a bytes-like object is required, not 'str'
/Users/armin/venv/weplan3.9/lib/python3.9/site-packages/dynaconf/base.py:809: TypeError
```
**Describe the solution you'd like**
I know that nowhere in the documentation the user is encouraged to pass `REDIS_FOR_DYNACONF`. But as it is possible after all, maybe the passed configuration should be patched to ensure `decode_responses` is set? | 0easy
|
Title: Direct Translation via URL request doesn't seem to work
Body: I wanted to directly get translations via the URL - e.g. load the translation with something like:
`https://libretranslate.com/?source=en&target=fr&q=I%2520love%2520it%21`
Unfortunately, this doesn't seem to work as only the source text is shown - even if I reload the page or press the "Translate Text" button, nothing happens. Only if I start editing the source text the translation shows up eventually.
I'm using Firefox 103 currently. | 0easy
|
Title: [SDK] Support Docker image as objective in the `tune` API
Body: Ref discussion: https://github.com/kubeflow/website/pull/3723#discussion_r1590261777.
Currently, user can only pass the training function as [objective in the `tune` API](https://github.com/kubeflow/katib/blob/af900202c67843650d80ff955e4e7fb7510cc68a/sdk/python/v1beta1/kubeflow/katib/api/katib_client.py#L156) in Katib Python SDK.
Similar to [`create_job` API in Training Python SDK](https://github.com/kubeflow/training-operator/blob/f23c5c8b0c87d93f61b539015883152a0872124f/sdk/python/kubeflow/training/api/training_client.py#L327), we should give user an ability to set objective as Docker image.
/area sdk
/good-first-issue
| 0easy
|
Title: Use of DataLoader to solve exorbitant query times
Body: There's a dramatic reduction in query performance when using GraphQL with graphene-mongo, as opposed to a regular REST API doing something like `list(collection.objects.all())` and sending the entire document set. From 300ms to over 6 seconds to retrieve one field from 100 elements.
I believe that using the DataLoader pattern is the expected solution to the N+1 problem, but actually attempting to use it will cause a runtime error
```
graphql.error.located_error.GraphQLLocatedError: Data loader batch_load_fn function raised an Exception: AttributeError("type object 'Person' has no attribute 'objects'",)
``` | 0easy
|
Title: Show better error messages when the exchange is not available
Body: In the Assignments and Courses extensions, if the exchange is not properly setup, the behavior is not very informative. The Assignments list just gives an error message, and the Courses extension just hangs. These should give better/more informative messages and point people to the documentation on setting up the exchange. | 0easy
|
Title: ENH: lazy init pandas fallbacks
Body: ### Is your feature request related to a problem? Please describe
To reduce the import time, we can init pandas fallback methods when users actually need them.
You can refer to the impl of [numpy fallbacks](https://github.com/xprobe-inc/xorbits/blob/0ad7c68d5517c2368653652ea14971a9dd2f6613/python/xorbits/numpy/random/__init__.py#L38).
You should also init pandas fallbacks when `__dir__` being called. | 0easy
|
Title: Add documentation related to how to handle `view_closed` events
Body: While there are docs for handling `view_submission` events (https://slack.dev/bolt-python/concepts#view_submissions), there are none for handling `view_closed` events.
In order to handle those events in your Bolt python app, two things need to be done:
- Add a `notify_on_close` property to your view payload when creating a view (see https://github.com/slackapi/bolt-python/issues/624)
- Use the `@app.view_closed('modal-callback-id')` decorator to define a view_closed handler
So we should document that! Can compare to the bolt-js docs for how bolt documents handling the view_closed event (bottom of https://slack.dev/bolt-js/concepts#view-submissions) | 0easy
|
Title: [DOC] Update paper referenced in Arsenal docs
Body: ### Describe the issue linked to the documentation
https://www.aeon-toolkit.org/en/stable/api_reference/auto_generated/aeon.classification.convolution_based.Arsenal.html#aeon.classification.convolution_based.Arsenal
references the arxiv paper rather than the open access paper in machine learning
https://link.springer.com/article/10.1007/s10994-021-06057-9
just a note to self and an excuse to remind everyone we have a classifier called Arsenal :)
### Suggest a potential alternative/fix
_No response_ | 0easy
|
Title: update docs to reflect new YB datasets module
Body: Update the documentation to leverage new yellowbrick.datasets & create documentation on metadata on example datasets such as number of samples, number of features, target/class.
For example, [sklearn.datasets.load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html).
Note: reminder to also update
- [x] the [corpus loading docs](https://www.scikit-yb.org/en/develop/api/text/corpus.html)
- [x] contributor guide with use of datasets/fixtures in tests (both random and real) | 0easy
|
Title: Using bare `Union` as annotation is not handled properly
Body: ```py
from typing import Union
def foo(a: Union): ...
```
```robot
Tc
foo 1
```
```
👉 python -m robot.run test.robot
==============================================================================
Test
==============================================================================
Tc | FAIL |
AttributeError: __args__
------------------------------------------------------------------------------
``` | 0easy
|
Title: Syntax error running jupyter-repo2docker
Body: I ran into a syntax error after installing `jupyter-repo2docker` in my conda environment.
I get the following when I try to do anything with jupyter-repo2docker.
```sh
admins-mbp-3:~ jacquelineburos$ jupyter-repo2docker --version
Traceback (most recent call last):
File "/Users/jacquelineburos/anaconda3/bin/jupyter-repo2docker", line 7, in <module>
from repo2docker.__main__ import main
File "/Users/jacquelineburos/anaconda3/lib/python3.5/site-packages/repo2docker/__main__.py", line 1, in <module>
from .app import Repo2Docker
File "/Users/jacquelineburos/anaconda3/lib/python3.5/site-packages/repo2docker/app.py", line 592
self.log.info(f'Successfully pushed {self.output_image_spec}', extra=dict(phase='pushing'))
^
SyntaxError: invalid syntax
```
Looks like your python_requires should be updated to >=3.6 since you are using f-strings.
Here is my system setup:
```py
Current conda install:
platform : osx-64
conda version : 4.3.22
conda is private : False
conda-env version : 4.3.22
conda-build version : 0+unknown
python version : 3.5.2.final.0
requests version : 2.14.2
root environment : /Users/jacquelineburos/anaconda3 (writable)
default environment : /Users/jacquelineburos/anaconda3
envs directories : /Users/jacquelineburos/anaconda3/envs
/Users/jacquelineburos/.conda/envs
package cache : /Users/jacquelineburos/anaconda3/pkgs
/Users/jacquelineburos/.conda/pkgs
channel URLs : https://repo.continuum.io/pkgs/free/osx-64
https://repo.continuum.io/pkgs/free/noarch
https://repo.continuum.io/pkgs/r/osx-64
https://repo.continuum.io/pkgs/r/noarch
https://repo.continuum.io/pkgs/pro/osx-64
https://repo.continuum.io/pkgs/pro/noarch
config file : /Users/jacquelineburos/.condarc
netrc file : None
offline mode : False
user-agent : conda/4.3.22 requests/2.14.2 CPython/3.5.2 Darwin/17.7.0 OSX/10.13.6
UID:GID : 502:20
``` | 0easy
|
Title: [BUG] sktime has codec error in Windows 11 Traditional Chinese System
Body: ```python
from sktime.forecasting.arima import AutoARIMA
```
cause UnicodeDecodeError as follwoing:
'''
File "C:\Users\lgzhangjian\scoop\apps\python310\3.10.11\lib\site-packages\sktime\base\_base.py", line 68
, in <module>
from sktime.utils._estimator_html_repr import _HTMLDocumentationLinkMixin
File "C:\Users\lgzhangjian\scoop\apps\python310\3.10.11\lib\site-packages\sktime\utils\_estimator_html_r
epr.py", line 190, in <module>
_STYLE = style_file.read()
UnicodeDecodeError: 'cp950' codec can't decode byte 0xe2 in position 4505: illegal multibyte sequence
'''
It is because "_estimator_html_repr.css" codec is 'utf8',
but "open" in Windows 11 Traditional Chinese System with default codec 'cp950' as following:
with open(Path(__file__).parent / "_estimator_html_repr.css") as style_file: # default codec with cp950 in windows 11 tranditional chinese
# use the style defined in the css file
_STYLE = style_file.read()
it should open with specified codec with 'utf8'.
with open(Path(__file__).parent / "_estimator_html_repr.css", encoding='utf8') as style_file: chinese
# use the style defined in the css file
_STYLE = style_file.read()
This fixes the problem.
sktime versions:
Name: sktime
Version: 0.34.0
| 0easy
|
Title: Web app manifest for assistant website
Body: I'd like to suggest that you add a web app manifest to the assistant website.
https://developer.mozilla.org/en-US/docs/Web/Manifest
https://web.dev/add-manifest/
With the correct settings (`"display": "standalone"`), this will enable users (using Chrome, Edge or Safari) to "Install" the assistant on mobile devices or desktop so that it appears nearly like a native app. | 0easy
|
Title: Revamp bot responses
Body: Rework responses for stuff like /gpt ask, and action menu items and etc to use embeds and look nicer. Generally improve aesthetic all around, would love some help on this. | 0easy
|
Title: Refactor preprocessing
Body: Refactor PreprocessingStep code. Right now there are `run` and `transform` methods, please provide `fit` method instead `run`. | 0easy
|
Title: dvc exp remove --keep
Body: I don't understand why one would want to _remove_ the [last n](https://dvc.org/doc/command-reference/exp/remove#-n) exps. Shouldn't this be in terms of `--keep` last `n`? The typical case is that one wants to just remove old exps.
| 0easy
|
Title: Show client's public ip address when "my ip" is used as search keyword
Body: The top search result should show client's public ip address when "My ip" keyword is used for search. Google does it.
This prevents the need to visit any 3rd party website just to check client's public ip address. Useful for troubleshooting vpn and throughput issues.
| 0easy
|
Title: unit tests not triggered on tags
Body: when we make a new release, we tag the commit. however, the unit tests are not triggered when this happens:
https://github.com/ploomber/ploomber/tree/0.20 | 0easy
|
Title: [Feature request] Add apply_to_images to PixelDropout
Body: | 0easy
|
Title: MaltParser is fixed to 1.7.2
Body: From #2802 (@tomaarsen)
The tests for MaltParser are fixed to version 1.7.2. Perhaps this means that users who have another version for personal use won't have passing tests? If so, that is a problem. Perhaps the best solution is to not even have to specify a version in the constructor. I'm not even sure it gets used - as I believe the MaltParser executable is actually loaded through the `MALT_PARSER` environment variable. | 0easy
|
Title: Change API spec file default name
Body: Current the default name for the API spec is `api.yaml`. One suggestions is to change it to something more common in the industry like `Scanfile` (similar to [Fastfile](https://docs.fastlane.tools/advanced/Fastfile/#fastfile) or [Dockerfile](https://docs.docker.com/engine/reference/builder/))
https://github.com/scanapi/scanapi/blob/master/scanapi/settings.py#L10
Besides we also need to:
- Update documentation at [scanapi/website](https://github.com/scanapi/website/tree/master/_docs_v1)
- Update examples at [scanapi/examples](https://github.com/scanapi/examples) | 0easy
|
Title: Support `times` and `x` suffixes with `WHILE` limit to make it more compatible with `Wait Until Keyword Succeeds`
Body: **Current Behaviour:**
Currently, WHILE loop have argument 'limit' which handles both 'timeout' and 'number of times loop to iterate'.
In most of the Robotframework keywords, timeout functionality supports values without suffix (for example, argument value 300 is equivalent to 300 seconds)
When developers use the limit argument for timeout, they forget the suffix (s, sec, min...), as most of the keywords in RobotFramework supports timeout in seconds without suffix.
For example, below keywords in Robot Framework supports without suffix, except WHILE loop
```
Sleep 300 # Sleeps for 300 seconds
Wait Until Keyword Succeeds 300 30 My Keyword args # Retries My Keyword for 300 second, every 30 seconds
WHILE ${status}==${TRUE} limit=300 # Loops 300 times. This is conflicting with rest other keyword timeout argument values.
```
**Feature Request (Either of Possibility 1 OR Possibility 2):**
**Possibility 1:**
Can there be a timeout argument in addition to limit argument to support timeout even without suffix. (supports timestring, number or integer)
This will avoid additional instructions to RobotFramework Developers, that to be careful when providing 'limit' argument to WHILE loop. Below is the feature request example.
Example:
```
WHILE ${status}==${TRUE} limit=5 # Loops 5 times
WHILE ${status}==${TRUE} timeout=300 # Loops only for 5 minutes or 300 seconds
```
_**OR**_
**Possibility 2:**
Can there be a suffix added to limit, something like 'x' or 'times' as in 'Wait Until Keyword Succeeds'. Same argument name 'limit' should work for both timeout and number of times. 'timeout' argument name not required in this case.
Example:
```
WHILE ${status}==${TRUE} limit=5x # Loops 5 times
WHILE ${status}==${TRUE} limit=10 times # Loops 10 times
WHILE ${status}==${TRUE} limit=300 # Timeout after 5 minutes or 300 seconds of looping
WHILE ${status}==${TRUE} limit=2m # Timeout after 2 minutes of looping
```
| 0easy
|
Title: 'data' from get_intraday function coming back as a dictionary, not as a dataframe
Body: This results in a dictionary:
```
data, meta_data = ts.get_intraday(symbol=symbol,interval='1min', outputsize='full')
logging.info(data)
```
And this results in an error:
`logging.info(data.head(2))` | 0easy
|
Title: Deprecate asyncio aliases in `falcon.util`
Body: I have soft-deprecated them in the docs in https://github.com/falconry/falcon/pull/2252.
Emit a deprecation warning in a 4.x release, remove in 5.0. Unless we need them again for, e.g., Trio support (?) | 0easy
|
Title: Connect Pagination dependency with Query primitive from FastAPI
Body: | 0easy
|
Title: CI: Update type checker (mypy) to match best practices
Body: Contributors: we need your help with this! Get started here: https://github.com/cleanlab/cleanlab/issues/407
- [ ] Add strict type annotations for all existing code.
- [ ] Switch over CI to use mypy in strict mode (adding [--strict](https://mypy.readthedocs.io/en/stable/command_line.html#cmdoption-mypy-strict) flag to existing call) once we have full coverage with type annotations.
When adding strict type annotations to existing code, you can see functions are currently lacking them by running:
```
mypy --strict --install-types --non-interactive cleanlab
```
Helpful discussion here about custom types we want to use/introduce as part of this effort: https://github.com/cleanlab/cleanlab/pull/317
| 0easy
|
Title: Custom Components
Body: Currently only the white listed set of components can be rendered, while we can and should extend that list, we should also allow custom components.
The idea would be to define a component like
```py
class Custom(BaseModel):
data: Any
class_name: _class_name.ClassName = None
sub_type: str = Field(serializeation_alias='subType')
type: typing.Literal['Custom'] = 'Custom'
```
`subType` let's implementing code do a nice switch thing when deciding how to render custom components.
The default implementation should just show `data` as JSON with a message saying "not implemented". | 0easy
|
Title: BaseConfig.multiindex_strict type hint
Body: The [type hint for BaseConfig.multiindex_strict](https://github.com/unionai-oss/pandera/blob/3ea4f2c4a3a4443c13dd9741d35bb7c8cbb930ac/pandera/typing/config.py#L39) is currently `multiindex_strict: bool = False`. However, because it allows "filter" as a value, I believe this type hint should be `Union[bool, str]`, similar to [BaseConfig.strict](https://github.com/unionai-oss/pandera/blob/3ea4f2c4a3a4443c13dd9741d35bb7c8cbb930ac/pandera/typing/config.py#L29). | 0easy
|
Title: Document set_cookie() name/value restrictions (strings are limited to US-ASCII)
Body: 
当我对cookie设置中文的时候，is_ascii_encodable会返回False。但是，cookie其实是可以设置中文的，不知道你们出于什么考虑，一定要求set_cookie的name和value是is_ascii。有人能跟我解释下吗？ | 0easy
|
Title: Stochastic Fast indicator
Body: Hi
I am using the latest version of Pandas Ta as a replacement for TA-Lib
I was using STOCHF - Stochastic Fast from TA-Lib (http://www.tadoc.org/indicator/STOCHF.htm)
I did not see it in Pandas-ta.
I it possible to "create" it using the parameters passed to the pandas-ta.stoch function?
Thanks
Amit | 0easy
|
Title: Add Trend Quality Indicator
Body: Trend Quality was created by David Sepiashvili and was unveiled in an issue of Stocks and Commodities magazine in 2004.
The basic idea behind it is that it is not enough to rate a market as "trending" or "not trending". We must elaborate more on the "strength" or "quality" of a trend so that we can make better trading decisions. In the original issue, the author created 2 indicators called the trend quality "q" and "b" indicator. I won't discuss the "b" indicator here because I think it is less useful and adopted less by the trading community.
The "q" indicator is quite useful because it returns values based on the quality of the trend. From the author:
_Readings in the range from +2 to +5, or from -2 to -5, can indicate moderate trending, and readings above Q=+5 or below Q=-5 indicate strong trending. Strong upward trending often leads to the security’s overvaluing, and strong downward trending often results in the security’s undervaluing._
I myself have noticed tq values go as high as 20 since in FX markets can seemingly trend forever.
Since it's unveiling tq has been rewritten for a few trading platforms, most notably mt4/5, but you can also find versions for it in ThinkOrSwim and TradingView. The version you see in most trading platforms is the "q" indicator and it is directional in nature.
The goat of MT4/5 indicator development mladen did a megathread on trend quality which you can find here: https://www.mql5.com/en/forum/181010.
I tried to port trend quality to python but I haven't produced anything usable yet. Here's what I've done so far: https://github.com/kbs-code/trend_quality_python/tree/master
In that repo, the file https://github.com/kbs-code/trend_quality_python/blob/master/reference/trend_quality_q1_directional_indicator.mq4 is a good reference for how the indicator should behave in mt4.
In the mql4 code, the indicator works on a for loop as it iterates over close prices. I tried to reverse engineer this behavior and apply the calculation in a vectorized manner using pandas data frames but failed. I've considered using lambdas or itertuples and have avoided trying iterrows because it's frowned upon. I've tried masks as well but I'm afraid that might introduce look-ahead bias.
Sources:
https://www.prorealcode.com/wp-content/uploads/2019/03/24-Q-Indicator.pdf
https://github.com/kbs-code/trend_quality_python/tree/master
https://www.mql5.com/en/forum/181010
https://usethinkscript.com/threads/trend-quality-indicator-with-alert-for-thinkorswim.9/
https://www.tradingview.com/script/k3mnaGcQ-Trend-Quality-Indicator/
| 0easy
|
Title: napari random sample data should have the seed set
Body: ## 🚀 Feature
I was trying to use napari's 3D balls set to illustrate something, and I found that the next run of the code returned a different set of balls. Although I can see the use of always having a random set, I think it's a bad experience for sample data and we should set the seed in the function signature ([here](https://github.com/napari/napari/blob/5535c71a787a5a17dc1ab8147e23b3bd0eff3131/napari_builtins/_ndims_balls.py#L6) and [here](https://github.com/napari/napari/blob/5535c71a787a5a17dc1ab8147e23b3bd0eff3131/napari_builtins/_ndims_balls.py#L19)), rather than choose a random seed within the function.
So the proposal is:
```python
def labeled_particles3d(shape=(256, 512, 512), seed=0):
labels, density, points = labeled_particles(
shape, seed=seed, return_density=True
)
...
```
(Might as well have shape in there too.) | 0easy
|
Title: [k8s] Add tini/other `init` for SkyPilot k8s pods
Body: [tini](https://github.com/krallin/tini) or [dumb-init](https://github.com/Yelp/dumb-init) act as PID 1 to collect zombie processes and handle signals from k8s. We should add them to our docker images + k8s manifests when we launch SkyPilot pods. | 0easy
|
Title: Invalid model identifier
Body: https://github.com/CTFd/CTFd/blob/master/CTFd/themes/core/templates/scoreboard.html#L26
This should change depending on the mode of the CTF | 0easy
|
Title: Improve warning message format
Body: The current warning messages via `warnings.warn` is hard to read due to the call trace that is printed alongside the warning. This is true even when we set the `stacklevel` parameter in `warnings.warn`. For example:
We should add warnings formatted that display nicer warning messages following best practices for warning formatting (refer to warning messages formatting in numpy and pandas?). This could possibly be achieved by a global levels setting or a warnings formatter.
| 0easy
|
Title: texture calc in wrong place.
Body: This is my bad for not catching it on the PR. But texture from @rcjackson is buried in gatefilters. Method should be exposed in pyart.retrieve
Needs to be moved. | 0easy
|
Title: Update documentation links.
Body: Many links in the documentation are 404 (for example) continuum is now anaconda, and make th test fails.
It should be easy to either update the links, or remove rephrase the corresponding sections. | 0easy
|
Title: [FEATURE] Spark and Dask Take 1 row without sorting optimization
Body: **Is your feature request related to a problem? Please describe.**
Look at [here](https://github.com/fugue-project/fugue/blob/838fdaa794c62e8bdc7f1474818d9491d5d39ed7/fugue_spark/execution_engine.py#L513)
If taking just one row with our sorting, we may use `GROUP BY` and `FIRST` to solve this problem, it can be a lot faster. Let's add this special handling. | 0easy
|
Title: Assign unique color to different values in unsorted bar charts
Body: Currently, all bars in the bar chart has the same color, this makes comparison across items challenging. In particular, users need to read the bar labels to realize that the bars are sorted to match other charts for comparison. We should assign a consistent set of colors for low cardinality categorical attributes to facilitate ease of comparison. We can use a standard color pallette for the particular visualization library ([Vega](https://vega.github.io/vega/docs/schemes/#categorical), [matplotlib](https://matplotlib.org/3.1.1/tutorials/colors/colormaps.html)).
The update bars would look something like this, where all values of Cylinder=4 is green, all values of Cylinders of 8 is yellow, etc.
| 0easy
|
Title: ODIM h5 file stays open after reading when using `pyart.aux_io.read_odim_h5`
Body: TL;DR: The function `pyart.aux_io.read_odim_h5` does not close the hdf5 files after reading the file, whether it read it successfully or raised an error. I have worked on a fix I'll send a pull request.
H5 files are kept open open in memory when read with the odim_h5 reader. Any attempt to read the file again fails and throws an OSError. The solution is to do a `gc.collect()`.
Moreover, if an error is raised while reading a file with `pyart.aux_io.read_odim_h5`, then that file can't be accessed again until you shutdown the python kernel or did a `gc.collect()`. To reproduce this error you can, for example:
- Open a cf/radial file with `pyart.aux_io.read_odim_h5`, it will throw an expected error: `KeyError: "Unable to open object (object 'what' doesn't exist)"`
- Try to open the same file with the correct function `pyart.io.read_cfradial`, you get:
```python
pyart/io/cfradial.py in read_cfradial(filename, field_names, additional_metadata, file_field_names, exclude_fields, include_fields, delay_field_loading, **kwargs)
125
126 # read the data
--> 127 ncobj = netCDF4.Dataset(filename)
128 ncvars = ncobj.variables
129
netCDF4/_netCDF4.pyx in netCDF4._netCDF4.Dataset.__init__()
netCDF4/_netCDF4.pyx in netCDF4._netCDF4._ensure_nc_success()
OSError: [Errno -101] NetCDF: HDF error
```
| 0easy
|
Title: CLI hiding the error traceback
Body: When a user operation triggers an error, we only print the error message if the exception is a `BaseException` (or a subclass):
https://github.com/ploomber/ploomber/blob/0a5fed2f9ec539c744664b1b7b3a7f744ec6c482/src/ploomber/cli/io.py#L99
However, when it's not a subclass, we want to show all the traceback, as it contains important details to debug. But the current implementation only prints the error message:
https://github.com/ploomber/ploomber/blob/0a5fed2f9ec539c744664b1b7b3a7f744ec6c482/src/ploomber/cli/io.py#L102
This makes debugging harder, see this https://github.com/ploomber/ploomber/issues/789#issuecomment-1130066043 | 0easy
|
Title: DOC: update sklearn documentation
Body: Add user guide and docs on xorbits.sklearn | 0easy
|
Title: Incorrect type hint for json= field of session.request when passing a list[dict]
Body: **Is your feature request related to a problem? Please describe.**
When I try to pass a list of dictionaries to session.request(), mypy's type checker shows an error
**Describe the solution you'd like**
JSON data may also be a list[dict]
**Describe alternatives you've considered**
—
**Additional context**
<img width="500" alt="Image" src="https://github.com/user-attachments/assets/06b6b4b6-a39d-48f9-a526-32cd5e3acf3f" />
| 0easy
|
Title: Migrate frontend tests to use Jest test style
Body: ## Description
We would like to align our tests and make this repository more ready for bun as it looks very promising.
The first step to do so is to align our current tests to use Jest style syntax only instead of both Chai and Jest syntax.
Basically all instances of:
```ts
expect(result).to.deep.eq()
```
Should be switched to:
```ts
expect(result).toEqual()
```
Both are supported by vitest so there should be no conflicts in doing this step by step. | 0easy
|
Title: Documentation incorrectly claims that `--tagdoc` documentation supports HTML formatting
Body: Hello,
Issue found with
$ rebot --version
Rebot 7.0.1 (Python 3.10.11 on win32)
I'm trying to use basic formatting with --tagdoc option, like
rebot --tagdoc "my_tag:\*my_doc\*" report.xml
But it's not interpreted like bold formatting like documentation said, same for italic formatting.
--tagdoc pattern:doc * Add documentation to tags matching the given
pattern. Documentation is shown in `Test Details` and
also as a tooltip in `Statistics by Tag`. Pattern can
use `*`, `?` and `[]` as wildcards like --test.
Documentation can contain formatting like --doc.
Examples: --tagdoc mytag:Example
--tagdoc "owner-*:Original author"
Also, --doc support documentation from file, but this is not the case for --tagdoc.
Rgds
SebC
| 0easy
|
Title: Specify `filename` and `content_type` headers through `Part` annotation
Body: **Is your feature request related to a problem? Please describe.**
Related to a problem surfaced by @Shanoir on [Gitter](https://gitter.im/python-uplink/Lobby?at=5d9f6234eac5612d22ec000f).
**Describe the solution you'd like**
I think that it would be a good idea for us to let users set the filename and content-type headers through the Part annotation. Maybe like so:
```python
@multipart
@retry(when=raises(Exception) | status(400),backoff=fixed(4),stop=after_attempt(3))
@headers(headersCollector)
@post("api/v1/data/file-upload")
def uploadFile(self, file: Part(filename="data.edl", content_type="application/octet-stream"), serialNumber: Field, append: Field): pass
```
| 0easy
|
Title: datetime locator doesn't work with cftime
Body: #### Code sample, a copy-pastable example if possible
```python
import proplot as plot
import numpy as np
import xarray as xr
times = xr.cftime_range('1990', '2000', freq='M')
data = xr.DataArray(np.random.rand(len(times)),
dims=['time'],
coords=[times])
f, ax = plot.subplots(aspect=2)
ax.plot(data.time, data)
# ax.format(xlocator='year')
# ax.format(xformatter='concise')
```
#### Actual result vs. expected result
The x datetime axis appears by default (with the comments included above), but when using `ax.format(xlocator='year')` or any locator, the xticks and labeling completely disappear.
When using any formatter (e.g. `xformatter='concise')` it breaks with the following error:
```python-traceback
ValueError: Cannot convert -3652 to a date. This often happens if non-datetime values are passed to an axis that expects datetime objects.
``` | 0easy
|
Title: Make listener v3 the default listener API
Body: Currently when creating listeners, you need to explicitly specify are you using listener API version 2 or 3. Now that the v3 API is getting better (#3296), we can make it the default. In other words, if API version isn't specified, we assume the v3 API is used. | 0easy
|
Title: Add unit tests for the `__repr__` method of the SpecEvaluator class
Body: ## Unit Test
### Description
Add unit tests for the `__repr__` method of the **SpecEvaluator** class: https://github.com/scanapi/scanapi/blob/main/scanapi/evaluators/spec_evaluator.py#L42
[ScanAPI Writing Tests Documentation](https://github.com/scanapi/scanapi/wiki/Writing-Tests) | 0easy
|
Title: Add correct article for the yeo-johnson transformation
Body: We need to add the following article as reference 1 [here](https://github.com/feature-engine/feature_engine/blob/main/feature_engine/transformation/yeojohnson.py#L72):
Yeo, In-Kwon and Johnson, Richard (2000). A new family of power transformations
to improve normality or symmetry. Biometrika, 87, 954-959.
And the current reference 1 should be reference 2. | 0easy
|
Title: param col_names for indicators with one resulting column
Body: Some indicators return multiple columns, then it is possible to rename the returned columns with `col_names` but for indicators with just one column this is not working e.g. for kind wma
The renaming happens at https://github.com/twopirllc/pandas-ta/blob/b2f2cc83a16376c1eb0a11aebed52186d7eab121/pandas_ta/core.py#L406
but i cant seee why it should not work for indicators with just one resulting column | 0easy
|
Title: Lots of dynamic notification banners when I invite a couple people on the doc
Body: ## Feature Request
**Is your feature request related to a problem or unsupported use case? Please describe.**
If I invite a few people on the doc, one notification banner appears for each user. Which is annoying
**Describe the solution you'd like**
Can we make a single notification for the whole group or just change the UX.
Or better if invitations went through, just close the pane and notify me if there is an error.
| 0easy
|
Title: DataArray.get_axis_num is incorrectly typed
Body: ### What happened?
When using mypy to check my code, I'm getting a spurious error. When passing a single string to `get_axis_num`, the return type is determined to be `tuple[int, ...]` instead of `int`.
### What did you expect to happen?
I expected the return type to be `int` when passing a single single.
### Minimal Complete Verifiable Example
```Python
# Note: to be run through mypy
import xarray as xr
def foo(x: xr.DataArray) -> None:
reveal_type(x.get_axis_num("time"))
```
### MVCE confirmation
- [X] Minimal example — the example is as focused as reasonably possible to demonstrate the underlying issue in xarray.
- [X] Complete example — the example is self-contained, including all data and the text of any traceback.
- [ ] Verifiable example — the example copy & pastes into an IPython prompt or [Binder notebook](https://mybinder.org/v2/gh/pydata/xarray/main?urlpath=lab/tree/doc/examples/blank_template.ipynb), returning the result.
- [X] New issue — a search of GitHub Issues suggests this is not a duplicate.
- [X] Recent environment — the issue occurs with the latest version of xarray and its dependencies.
### Relevant log output
```Python
# Output from mypy
test.py:4: note: Revealed type is "builtins.tuple[builtins.int, ...]"
Success: no issues found in 1 source file
```
### Anything else we need to know?
From a quick look at the code, I'm guessing this is because `str` is iterable and so is matching the first `@overload`.
### Environment
<details>
INSTALLED VERSIONS
------------------
commit: None
python: 3.12.3 (main, Nov 6 2024, 18:32:19) [GCC 13.2.0]
python-bits: 64
OS: Linux
OS-release: 6.8.0-49-generic
machine: x86_64
processor: x86_64
byteorder: little
LC_ALL: None
LANG: C.UTF-8
LOCALE: ('C', 'UTF-8')
libhdf5: 1.14.2
libnetcdf: None
xarray: 2024.11.0
pandas: 2.1.4
numpy: 1.26.2
scipy: 1.11.4
netCDF4: None
pydap: None
h5netcdf: None
h5py: 3.11.0
zarr: None
cftime: None
nc_time_axis: None
iris: None
bottleneck: None
dask: 2023.12.0
distributed: None
matplotlib: 3.9.1
cartopy: None
seaborn: None
numbagg: None
fsspec: 2023.12.1
cupy: 13.2.0
pint: None
sparse: None
flox: None
numpy_groupies: None
setuptools: 75.1.0
pip: 24.3.1
conda: None
pytest: 8.2.2
mypy: 1.11.1
IPython: 8.18.1
sphinx: 7.2.6
</details>
| 0easy
|
Title: Regression metric to be optimized
Body: RF is optimizing MSE but we report in the print the RMSE. Please verify this. | 0easy
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.