Changed import scheme for compliance with pip install (#46)

* Added arg to specify type of elements for correlogram

* More verbose errors for wrong column selection in log transform

* Improved parquet_to_pg to enable custom column selection before aggregation

* Included pandera in pyproject.toml

* Refactored general imports and preprocessing

* Refactored visualization

* Refactored analysis

* Black formatting and import optimisation

* Next try on black formatting

* Fixed typo in pyproject.toml

* Updated pyproject.toml to include submodules in installation

* Updated README.md

Author: aretaon

README.md

@@ -26,7 +26,6 @@ Its main features are:
 
 - Generate a new python environment using [anaconda](https://conda.io/projects/conda/en/latest/user-guide/tasks/manage-environments.html)
 or [pip](https://packaging.python.org/en/latest/guides/installing-using-pip-and-virtual-environments/).
-  - Install required Python packages (see [pyproject.toml](pyproject.toml)) in the environment
 - Download or clone autoprot
   - If you clone the repository, make sure to include the dependencies as submodules (example below)
 
@@ -40,7 +39,23 @@ git clone --recurse-submodules https://github.com/ag-warscheid/autoprot.git
 git submodule update --init --recursive
 ```
 
-- Next you need to install R. Please follow the instructions at the [R manual](https://cran.r-project.org/index.html) and install R to a custom location
+- Install autoprot to your Python env using
+
+```
+cd autoprot
+pip install .
+```
+- Next you need to install R. Please follow the instructions from the [R manual](https://cran.r-project.org/index.html) and install R to a custom location
+  - On Linux you could typically do
+
+```bash
+wget https://cran.r-project.org/src/base/R-4/R-4.5.3.tar.gz
+tar -xzf R-4.5.3.tar.gz
+cd R-4.5.3
+./configure
+make
+```
+
 - Start autoprot by importing it from any Python console you like. It will generate an autoprot.conf file in the autoprot package directory that you need to edit.
   - Insert the path to your Rscript executable that you just installed as value for the R variable
   - The RFunctions variable should point the RFunctions.R file from autoprot.

autoprot/__init__.py

@@ -1,5 +1,8 @@
 import sys
+
 from . import common
+from . import decorators
+from . import r_helper
 
 __version__ = "dev"
 __author__ = "The autoprot contributors"
@@ -9,6 +12,7 @@
 module_pointer = sys.modules[__name__]
 # we can explicitly make assignments on it
 module_pointer.check_r_install = False
-
 # write the environment.txt file
 common.generate_environment_txt()
+
+from . import preprocessing

autoprot/analysis/PCA.py

@@ -9,22 +9,23 @@
 
 from typing import Union, Literal
 
-import pandas as pd
-import numpy as np
 import matplotlib.pylab as plt
+import numpy as np
+import pandas as pd
 import seaborn as sns
-
+from gprofiler import GProfiler
 from sklearn.decomposition import PCA
-from .. import r_helper
 
-from gprofiler import GProfiler
+from autoprot import r_helper
 
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = ["AutoPCA"]
+
 
 class AutoPCA:
     # noinspection PyUnresolvedReferences

autoprot/analysis/__init__.py

@@ -1,6 +1,16 @@
+from .PCA import *
 from .clustering import *
 from .functional import *
-from .PCA import *
 from .qc import *
 from .stat_test import *
 from .stats import *
+
+# Optional: define what's exported when someone does `from analysis import *`
+__all__ = [
+    "clustering",
+    "functional",
+    "PCA",
+    "qc",
+    "stat_test",
+    "stats",
+]

autoprot/analysis/clustering.py

@@ -16,10 +16,10 @@
 import matplotlib.pylab as plt
 import numpy as np
 import pandas as pd
-from pandas.api.types import is_numeric_dtype
 import seaborn as sns
 from gprofiler import GProfiler
 from numpy.typing import ArrayLike
+from pandas.api.types import is_numeric_dtype
 from scipy import cluster as clst
 from scipy.spatial import distance
 from scipy.stats import zscore
@@ -30,14 +30,16 @@
     davies_bouldin_score,
 )
 
-from .. import r_helper
+from autoprot import r_helper
 
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = ["HCA", "KMeans"]
+
 
 class _Cluster:
     r"""

autoprot/analysis/functional.py

@@ -11,22 +11,23 @@
 from importlib import resources
 from typing import Union, Literal, List, Dict, Any, Tuple
 
-import pandas as pd
-import numpy as np
 import matplotlib.pylab as plt
+import numpy as np
+import pandas as pd
 import seaborn as sns
-
-from .. import visualization as vis
-from .. import r_helper
-
 from gprofiler import GProfiler
 
+from autoprot import r_helper
+from autoprot import visualization as vis
+
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = ["go_analysis", "KSEA"]
+
 
 def go_analysis(
     gene_list: list[str],

autoprot/analysis/qc.py

@@ -7,26 +7,34 @@
 @documentation: Julian
 """
 
-from typing import Literal
 from datetime import date
+from operator import itemgetter
+from typing import Literal
 
-import pandas as pd
-import numpy as np
 import matplotlib.pylab as plt
-import seaborn as sns
-from operator import itemgetter
 import missingno as msn
-
-from .. import r_helper
-
+import numpy as np
+import pandas as pd
+import seaborn as sns
 from gprofiler import GProfiler
 
+from autoprot import r_helper
+
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = [
+    "miss_analysis",
+    "missed_cleavages",
+    "enrichment_specificity",
+    "SILAC_labeling_efficiency",
+    "dimethyl_labeling_efficieny",
+    "tmt6plex_labeling_efficiency",
+]
+
 
 def miss_analysis(
     df,

autoprot/analysis/stat_test.py

@@ -17,15 +17,17 @@
 from scipy.stats import ttest_1samp, ttest_ind
 from statsmodels.stats import multitest as mt
 
-from .. import preprocessing as pp
-from .. import r_helper
+from autoprot import preprocessing as pp
+from autoprot import r_helper
 
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = ["ttest", "adjust_p", "cohen_d", "limma", "rank_prod"]
+
 
 def ttest(
     df,

autoprot/analysis/stats.py

@@ -7,21 +7,22 @@
 @documentation: Julian
 """
 
-import pandas as pd
-import numpy as np
 import matplotlib.pylab as plt
+import numpy as np
+import pandas as pd
 import seaborn as sns
-
-from .. import r_helper
-
 from gprofiler import GProfiler
 
+from autoprot import r_helper
+
 gp = GProfiler(user_agent="autoprot", return_dataframe=True)
 RFUNCTIONS, R = r_helper.return_r_path()
 
 # check where this is actually used and make it local
 cmap = sns.diverging_palette(150, 275, s=80, l=55, n=9)
 
+__all__ = ["edm", "loess", "make_psm"]
+
 
 def edm(matrix_a, matrix_b):
     """