Shaokai/fix install

README.md

@@ -25,16 +25,28 @@
 
 ## Install & Run AmadeusGPT🎻
 
-- AmadeusGPT is a Python package hosted on `pypi`. You can create a virtual env (conda, etc, see below*) or Docker and run:
-```python
-pip install 'amadeusgpt[streamlit]'
+- We prepare installation script install.sh. Make sure you modify the bash script to fit your own.
+```bash
+#!/bin/bash
+# replace this with your own conda path
+source /Users/shaokaiye/miniforge3/bin/activate
+conda env create -f conda/amadesuGPT.yml
+conda activate amadeusgpt
+pip install -e .[streamlit]
 ```
+- After the installation, you can do 
+
+```bash
+conda activate amadeusgpt
+```
+
  - Please note that you need an [openAI API key](https://platform.openai.com/account/api-keys), which you can easily create [here](https://platform.openai.com/account/api-keys).
 - If you want the **Streamlit Demo on your computer**, you will also need demo files that are supplied in our repo (see below**), so please git clone the repo and navigate into the `AmadeusGPT` directory. Then in your conda env/terminal run  `pip install 'amadeusgpt[streamlit]'` as described above. Then, to launch the Demo App execute in the terminal:
-```python
+```bash
 make app
 ```
-- You can use AmadeusGPT directly in python: iPython, Jupyter Notebooks, Google Colab, etc. For a quick start, see our `\examples` directory that hosts demo data and a Jupyter Notebook! Enjoy!
+- You can use AmadeusGPT directly in python: iPython, Jupyter Notebooks etc. For a quick start, see our `\examples` directory that hosts demo data and a Jupyter Notebook! Enjoy!
+-  Note if you use Google Colab, you still need to clone the repo (to access the demo data and installation script) and run installation script (after modifying the conda path)
 
 ## Citation
 
@@ -51,12 +63,10 @@ make app
 - arXiv preprint version **[AmadeusGPT: a natural language interface for interactive animal behavioral analysis](https://arxiv.org/abs/2307.04858)** by [Shaokai Ye](https://github.com/yeshaokai), [Jessy Lauer](https://github.com/jeylau), [Mu Zhou](https://github.com/zhoumu53), [Alexander Mathis](https://github.com/AlexEMG) & [Mackenzie W. Mathis](https://github.com/MMathisLab).
 
 ### Install tips
-- *Make a new conda env: `conda create --name amadeusGPT python=3.9` then run `conda activate amadeusGPT` or you can also use our supplied conda if you git cloned the repo (navigate into the conda directory): `conda env create -f amadesuGPT.yml` then pip install amadeusGPT once created/launched.
 - **Git clone this repo: so please open a terminal, we recommend to download into Documents (so type `cd Documents`) and run `git clone https://github.com/AdaptiveMotorControlLab/AmadeusGPT.git` Then go into the dir (`cd AmadeusGPT`)
 - If you want to use SAM, you need to download the weights. Otherwise you will see the following message in the app: `Cannot find SAM checkpoints. Skipping SAM`. Download them and add to "static" directory: wget https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth
 
 ###  Install trouble shooting:
-- If you hit an error during installing on an M1/M2 Macbook with installing HDF5, run `conda install h5py` in your conda env.
 - If you launch the app and get an ffmpeg error, `RuntimeError: No ffmpeg exe could be found. Install ffmpeg on your system, or set the IMAGEIO_FFMPEG_EXE environment variable.` try running `conda install ffmpeg`.
 - If you have an M1/M2 chip and use CEBRA within AmadeusGPT, and you get this error: `RuntimeError: Device type MPS is not supported for torch.Generator() api` run `pip install --upgrade torch`.
 

amadeusgpt/implementation.py

@@ -22,8 +22,6 @@
 import cv2
 import matplotlib.path as mpath
 import msgpack
-from moviepy.video.io.ffmpeg_writer import FFMPEG_VideoWriter
-from moviepy.video.io.VideoFileClip import VideoFileClip
 from PIL import Image
 from pycocotools import mask as mask_decoder
 from scipy.signal import savgol_filter
@@ -1404,26 +1402,21 @@ def get_kinematics(self, bodyparts: List[str], kin_type: str):
         """
         assert kin_type in ["location", "velocity", "acceleration", "speed"]
         ret = None
-        import dlc2kinematics
+
 
         df = Database.get(type(self).__name__, "df")
         n_kpts = Database.get(type(self).__name__, "n_kpts")
         n_individuals = Database.get(type(self).__name__, "n_individuals")
 
         if kin_type == "velocity":
-            ret = dlc2kinematics.compute_velocity(df, bodyparts=bodyparts)
+            ret = AnimalBehaviorAnalysis.get_velocity()
 
         elif kin_type == "acceleration":
-            ret = dlc2kinematics.compute_acceleration(df, bodyparts=bodyparts)
+            ret = AnimalBehaviorAnalysis.get_acceleration()
 
         elif kin_type == "speed":
-            ret = dlc2kinematics.compute_speed(df, bodyparts=bodyparts)
-        elif kin_type == "location":
-            if bodyparts[0] == "all":
-                mask = np.ones(df.shape[1], dtype=bool)
-            else:
-                mask = df.columns.get_level_values("bodyparts").isin(bodyparts)
-            ret = df.loc[:, mask]
+            ret = AnimalBehaviorAnalysis.get_speed()
+
         else:
             raise ValueError(f"{kin_type} is not supported")
         n_kpts = len(bodyparts) if bodyparts != ["all"] else n_kpts
@@ -2794,30 +2787,4 @@ def animals_social_events(
 
         return AnimalAnimalEvent(ret_events)
 
-    def generate_videos_by_events(self, events: List[Event]):
-        """
-        Examples
-        --------
-        >>> # find where the animal is to the left of object 6 and create videos for those events
-        >>> def task_program():
-        >>>     behavior_analysis = AnimalBehaviorAnalysis()
-        >>>     left_to_object_events = behavior_analysis.animals_object_events('6', ['to_left'])
-        >>>     behavior_analysis.generate_videos_by_events(left_to_object_events)
-        >>>     return
-        """
-        video_file_path = AnimalBehaviorAnalysis.get_video_file_path()
-        clip = VideoFileClip(video_file_path)
-        fps = clip.fps
-        for event_id, event in enumerate(events):
-            start_frame = np.where(event.mask)[0][0]
-            end_frame = np.where(event.mask)[0][-1]
-            start_in_seconds = start_frame / fps
-            end_in_seconds = end_frame / fps
-            output_file = f"event{event_id}.mp4"
-            trimmed_clip = clip.subclip(start_in_seconds, end_in_seconds)
-            writer = FFMPEG_VideoWriter(
-                output_file, trimmed_clip.size, fps=trimmed_clip.fps
-            )
-            trimmed_clip.write_videofile(output_file)
-            print(f"generated video at {output_file}")
-        return None
+

amadeusgpt/logger.py

@@ -1,4 +1,4 @@
-from google.cloud import firestore
+
 import os
 import streamlit as st
 import psutil
@@ -16,10 +16,6 @@ class AmadeusLogger:
     logging.basicConfig(level=level)
     filename = None
     logger = logging.getLogger()
-    if "streamlit_cloud" in os.environ:
-        db = firestore.Client.from_service_account_json(
-            "amadeusgpt-a1849-firebase-adminsdk-eyfx1-4a1bbc8488.json"
-        )
 
     @classmethod
     def format(cls, message):

amadeusgpt/main.py

@@ -45,18 +45,6 @@
 
 ############
 # all these are for safe execution of generated code
-from amadeusgpt.amadeus_security import check_code
-from RestrictedPython import (
-    compile_restricted,
-    safe_globals,
-    safe_builtins,
-    utility_builtins,
-    limited_builtins,
-)
-from RestrictedPython.Guards import (
-    guarded_unpack_sequence,
-    guarded_iter_unpack_sequence,
-)
 
 ############
 

amadeusgpt/modules/implementation/embedding/umap.py

@@ -1,6 +1,5 @@
 import numpy as np
-import umap
-from umap.umap_ import UMAP
+import umap.umap_ as umap
 from amadeusgpt.implementation import AnimalBehaviorAnalysis
 from .transform import align_poses
 
@@ -10,7 +9,7 @@ def compute_embedding_with_umap_and_plot_embedding(
 ):
     features = inputs.reshape(inputs.shape[0], -1)
     features = np.nan_to_num(features)
-    reducer = UMAP(n_components=n_dimension, min_dist=0.5)
+    reducer = umap.UMAP()(n_components=n_dimension, min_dist=0.5)
     embedding = reducer.fit_transform(features)
 
     behavior_analysis = AnimalBehaviorAnalysis()

conda/amadesuGPT.yml

@@ -6,5 +6,5 @@ channels:
 dependencies:
     - python<3.10
     - pip
-    - pytorch
     - jupyter
+    - hdf5