Closed Kammani closed 2 weeks ago
Currently, cython and thus this extension is not compatible with numpy>=2.
You can restrict numpy<2
Try numpy v1.26.4. This solved the problem for my venv.
I had the same issue on my Mac Air M2-chip "Try numpy v1.26.4. This solved the problem for my venv." This fixed my same problem.
I made a new venv to keep numpy 1.x separate from 2.x for my other projects. After having done 'brew install ta-lib' and 'pip install TA-Lib' I did the following:
pip unintall numpy pip cache purge (this line may not be necessary...) pip install numpy==1.26.4
hope it's not a big trouble but can we update the dependency of numpy from "numpy *" to "numpy ^1.26.4" that'll help most people understand what's going on without asking.
What change do you want?On Nov 2, 2024, at 4:39 PM, QiuxiaoMu @.**> wrote: hope it's not a big trouble but can we update the dependency of numpy from "numpy " to "numpy ^1.26.4" that'll help most people understand what's going on without asking.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: @.***>
What change do you want?On Nov 2, 2024, at 4:39 PM, QiuxiaoMu @.> wrote: hope it's not a big trouble but can we update the dependency of numpy from "numpy " to "numpy ^1.26.4" that'll help most people understand what's going on without asking. —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you commented.Message ID: **@.***>
I might be wrong but with poetry's show
command normally we'd see the dependencies specified by the package itself and my first impression was that it was in the hand of the maintainer, not peotry.
Here's what I see from poetry show ta-lib
:
name : ta-lib
version : 0.4.32
description : Python wrapper for TA-Lib
dependencies
- numpy *
Here's a little to no help one, as there's not even a dependency section, from poetry show polars
:
name : pandas
version : 2.2.3
description : Powerful data structures for data analysis, time series, and statistics
And here's a very helpful one, giving us specific version for each dependency, from poetry show pandas
:
name : pandas
version : 2.2.3
description : Powerful data structures for data analysis, time series, and statistics
dependencies
- numpy >=1.26.0
- python-dateutil >=2.8.2
- pytz >=2020.1
- tzdata >=2022.7
required by
- alpaca-py >=1.5.3
- backtesting >0.25.0
- bokeh >=1.2
- exchange-calendars >=1.5
- nautilus_trader >=2.2.3,<3.0.0
- pandas-datareader >=0.23
- pandas-market-calendars >=1.1
- pyti *
- seaborn >=1.2
- statsmodels >=1.4,<2.1.0 || >2.1.0
I am not saying I wanted the change, I was just thinking at least for people using poetry it'll be easier for them to find out that ta-lib has a specific dependency on numpy and they don't have raise new issue just for the sake of duplication.
That’s odd because I thought it was specified as numpy<2 everywhere.
Released ta-lib 0.5.0 with numpy 2.0 support.
Also released ta-lib 0.4.33 with numpy<2.0 support.
numpy-2.0.1 and ta-lib-0.4.32 are used . And tests.py is simple code as mentioned here .
import numpy as np import talib
Check the versions to confirm the installation
print("NumPy version:", np.version) print("TA-Lib version:", talib.version)
Example usage of TA-Lib
data = np.random.random(100) # Generate some random data sma = talib.SMA(data, timeperiod=10) # Calculate a simple moving average
print("Simple Moving Average:", sma)
And the following message and Error appeared upon execution of tests.py file . ————— Traceback (most recent call last): File "/Users/Documents//testv.py", line 2, in
import talib
File "/Users//Documents//venv/lib/python3.12/site-packages/talib/init.py", line 93, in
from ._ta_lib import (
File "talib/_ta_lib.pyx", line 1, in init talib._ta_lib
ValueError: numpy.dtype size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject
——————-
Can you offer any solution ? Many thanks .