1

CTAPT.py

@@ -4,8 +4,11 @@ import streamlit as st, pandas as pd, numpy as np
 st.set_page_config(page_title="Предсказание V", page_icon="🦋", layout="wide", initial_sidebar_state="expanded")
 # set_page_config() can only be called once per app page, and must be called as the first Streamlit command in your script.
 
-import my_static_methods as my_stm
-st.html(my_stm.STYLE_CORRECTION)
+import my_static_tools as mst
+from my_static_tools import HF_tools as hft
+from my_static_tools import XYZV_tools as xvt
+
+st.html(mst.STYLE_CORRECTION)
 
 st.sidebar.markdown("💎 Стартовая страница")
 

my_static_methods.py

@@ -1,195 +0,0 @@
-from typing import Union, NamedTuple
-import io,os,re,sys,math,time,uuid,ctypes,pickle,random,shutil,string,urllib,decimal,datetime,itertools,traceback,collections,statistics
-import numpy as np, pandas as pd
-import plotly.express as px
-import huggingface_hub
-
-import sklearn #, statsmodels
-from sklearn import svm, neighbors, naive_bayes, neural_network, tree, ensemble, linear_model, discriminant_analysis, gaussian_process, manifold, cluster
-#from statsmodels.tsa import seasonal
-
-os.makedirs(".temp", exist_ok=True) # for temporary local files
-
-class HfRepo(NamedTuple):
-    repo_id: str
-    repo_type: str
-    token: str
-
-
-### remove decoration and popup menu button at top
-STYLE_CORRECTION = " ".join([
-    "<style>",
-    "header[data-testid='stHeader'] { display:none }",
-    "div[data-testid='stSidebarHeader'] { display:none }",
-    "div[data-testid='stAppViewBlockContainer'] { padding:1em }",
-    "div[data-testid='collapsedControl'] { background-color:#EEE }",
-    "a[href='https://streamlit.io/cloud'] { display:none }"
-    "</style>"
-])
-
-###
-def pandas_info(df: pd.DataFrame) -> Union[pd.DataFrame,str]:
-    buffer = io.StringIO()
-    df.info(buf=buffer)
-    str_info = buffer.getvalue()
-    try:
-        lines = str_info.splitlines()
-        df = (pd.DataFrame([x.split() for x in lines[5:-2]], columns=lines[3].split()).drop('Count',axis=1).rename(columns={'Non-Null':'Non-Null Count'}))
-        return df
-    except Exception as ex:
-        print(ex)
-        return str_info
-
-### случайные числа, для отладки например
-def df_random_dataframe(n_cols:int = 15, n_rows:int = 100) -> pd.DataFrame:
-    df = pd.DataFrame(np.random.randn(n_rows, n_cols), columns=(f"col {i}" for i in range(n_cols)))
-    return df
-
-### обработка столбца V для дальнейшего удобства + столб T типа время
-def df_process_v_column(df: pd.DataFrame) -> pd.DataFrame:
-    df = df.reset_index() #
-    df.rename(columns = {"index": "T"}, inplace=True)
-    df["Vis"] = df.V.map(lambda v: 0 if str(v)=="nan" else 1).astype(int)
-    df["Vfloat"] = df.V.map(lambda v: 0 if str(v)=="nan" else str(v).replace(',', '.')).astype(float)
-    df["Vsign"] = df.Vfloat.map(lambda v: -1 if v<0 else 1 if v>0 else 0).astype(int)
-    df["Vposneg"] = df.Vfloat.map(lambda v: "n" if v<0 else "p" if v>0 else "o").astype(str)
-    return df
-
-###
-def save_dataframe_to_hf(repo: HfRepo, dfToSave: pd.DataFrame, new_filename: str, remote_subdir: str) -> Union[huggingface_hub.CommitInfo, Exception]:
-    """ save dataframe to hf repo """
-    try:
-        local_filename = os.path.join(".temp", new_filename)
-        #df.to_csv('compressed_data.zip', index=False, compression={'method': 'zip', 'archive_name': 'data.csv'})
-        dfToSave.to_csv(local_filename, index=False, sep=";", encoding="utf-8") # , compression="zip"
-        apiHF = huggingface_hub.HfApi(token=repo.token)
-        path_in_repo = os.path.basename(local_filename)
-        if remote_subdir:
-            path_in_repo = f"{remote_subdir}/{path_in_repo}"
-        commit_info = apiHF.upload_file(path_or_fileobj=local_filename, path_in_repo=path_in_repo, repo_id=repo.repo_id, repo_type=repo.repo_type)
-        return commit_info
-    except Exception as exSave:
-        return exSave
-
-
-###
-def load_dataframes_from_hf(repo: HfRepo, lstCsvFiles: list[str] = []) -> {str, pd.DataFrame}:
-    """ load dataframes from hf """
-    #https://huggingface.co/datasets/f64k/gaziev/blob/main/TestData3_2204_noAB_gaziev.zip
-    dict_res = {}
-    for fl_name in lstCsvFiles:
-        try: file_loaded = huggingface_hub.hf_hub_download(filename=fl_name, repo_id=repo.repo_id, repo_type=repo.repo_type, token=repo.token)
-        except: file_loaded = ""
-        if os.path.exists(file_loaded):
-            compress = "zip" if file_loaded.lower().endswith("zip") else None
-            df_loaded = pd.read_csv(file_loaded, sep=";", encoding = "utf-8", compression=compress)
-            dict_res[fl_name] = df_loaded # df_Vproc = df_process_v_column(df_loaded)
-    return dict_res
-
-### список CSV и ZIP файлов (c уровнем вложенности) в репозитории
-### https://huggingface.co/docs/huggingface_hub/en/guides/hf_file_system
-def list_files_hf(repo: HfRepo) -> list[str]:
-    """ List CSV and ZIP files in HF repo """
-    fs = huggingface_hub.HfFileSystem(token=repo.token, use_listings_cache=False) # , skip_instance_cache=True
-    path_hf = f"{repo.repo_type}s/{repo.repo_id}/"
-    #lst = fs.ls(path_hf, detail=False)
-    lstGlob = fs.glob(path_hf + "**") # map(os.path.basename, lstGlob)
-    lstNames = [fname.replace(path_hf, "") for fname in lstGlob if fname.lower().endswith(".csv") or fname.lower().endswith(".zip")]
-    #print(f"ПРОЧИТАНО В list_files_hf() : {lstNames=}")
-    return lstNames
-
-###
-def plotly_xyzv_scatter_gray(df3D):
-    """ 3D plot """
-    color_discrete_map = dict(o='rgb(230,230,230)', p='rgb(90,1,1)', n='rgb(1,1,90)')
-    fig = px.scatter_3d(df3D, x='X', y='Y', z='Z', color="Vposneg", opacity=0.4, height=800, color_discrete_map=color_discrete_map)
-    fig.update_scenes(
-        xaxis={"gridcolor":"rgba(30, 0, 0, 0.2)","color":"rgb(100, 0, 0)","showbackground":False},
-        yaxis={"gridcolor":"rgba(0, 30, 0, 0.2)","color":"rgb(0, 100, 0)","showbackground":False},
-        zaxis={"gridcolor":"rgba(0, 0, 30, 0.2)","color":"rgb(0, 0, 100)","showbackground":False})
-    fig.update_traces(marker_size=3)
-    return fig
-
-
-# lstRepoZipFiles = ["TrainData_1504_AB_gaziev.zip","TestData_1504_AB_gaziev.zip","TestData3_2204_noAB_gaziev.zip"]
-### returns (classifier_object, df_train_with_predict, time_elapsed)
-def GetClassifier(lstDfOriginal, nHystorySteps) :
-    #lstDfOriginal = [df_9125_Train, df_12010_Test, df_9051_Test3]
-    nShift = nHystorySteps
-    nCurrShift = nHystorySteps
-    classifierName = "DecisionTreeClassifier"
-    colsVectorInp = ["X","Y","Z"]
-    fieldY = "Vis" #
-    lstDataFrames = MakeHystoryColumns(lstDfOriginal, nShift)
-    df_train = pd.concat(lstDataFrames)
-    lstColsShift = [f"{c}-{i}" for i in range(1, nCurrShift+1) for c in colsVectorInp] # для nCurrShift=0 lstColsShift=[]
-    colsVectorInpAll = colsVectorInp + lstColsShift
-    y_train = df_train[fieldY]
-    x_train_vect = df_train[colsVectorInpAll]
-    dictClassifiers = createDictClassifiers_BestForXYZ()
-    classifierObject = dictClassifiers[classifierName]
-    start2 = time.time()
-    classifierObject.fit(x_train_vect, y_train) # процесс обучения
-    time_elapsed = time.time() - start2
-    y_pred = classifierObject.predict(x_train_vect.values)  # .values[:,::-1] поля XYZ и истории в обратном порядке
-    df_train[f"predict_{fieldY}"] = y_pred
-    return (classifierObject, df_train, time_elapsed)
-
-#
-def MakeHystoryColumns(lstDfOriginal, nShift) :
-    lstDataframesShifted = [df.copy() for df in lstDfOriginal]
-    lstColsShift = []
-    for i in range(1, nShift+1):
-        #cols = ["X","Y","Z"]+["A","B"]
-        cols = ["X","Y","Z"]
-        #cols = ["A","B"]
-        for c in cols:
-            for dfShift in lstDataframesShifted:
-                dfShift[f'{c}-{i}'] = dfShift[c].shift(i).fillna(0)
-            lstColsShift.append(lstDataframesShifted[0].columns[-1])
-    print(lstColsShift)
-    return lstDataframesShifted
-
-RANDOM_STATE=11
-
-def createDictClassifiers_BestForXYZ() :
-    dictFastTree = {
-        #"RandomForestClassifier": ensemble.RandomForestClassifier(random_state=RANDOM_STATE), # совсем плохие показатели
-        #"ExtraTreeClassifier": tree.ExtraTreeClassifier(random_state=RANDOM_STATE), #
-        "DecisionTreeClassifier": tree.DecisionTreeClassifier(random_state=RANDOM_STATE), # лучший по последним баллам
-    }
-    #return {**dictFast}
-    #return {**dict_Test_MLPClassifier}
-    #return {**dictFast, **dictLongTrain}
-    return {**dictFastTree}
-
-
-
-
-#import joblib
-#REPO_ID = "YOUR_REPO_ID"
-#FILENAME = "sklearn_model.joblib"
-#model = joblib.load(hf_hub_download(repo_id=REPO_ID, filename=FILENAME))
-
-
-
-
-if False:
-    if False:
-        # https://scikit-learn.org/stable/auto_examples/preprocessing/plot_all_scaling.html
-        scaler = sklearn.preprocessing.StandardScaler()
-        #scaler = sklearn.preprocessing.PowerTransformer()
-        #scaler = sklearn.preprocessing.RobustScaler()
-        #scaler = sklearn.preprocessing.MinMaxScaler() # https://scikit-learn.org/1.1/modules/generated/sklearn.preprocessing.MinMaxScaler.html#sklearn.preprocessing.MinMaxScaler
-        #scaler = sklearn.preprocessing.MinMaxScaler(feature_range=(-1,1))
-        #scaler = sklearn.preprocessing.QuantileTransformer()
-        #scaler = sklearn.preprocessing.QuantileTransformer(output_distribution="normal")
-        #scaler = sklearn.preprocessing.Normalizer() # всё на сферу кладёт - приводит к 1 длину вектора
-        scale_columns = ["X","Y","Z"]
-        scaledData = scaler.fit_transform(df3D[scale_columns])
-        if False:
-            scaler2 = sklearn.preprocessing.Normalizer()
-            scaledData = scaler2.fit_transform(scaledData)
-        df3D_Scaled = pd.DataFrame(data=scaledData, columns=scale_columns)
-        df3D_Scaled["Vposneg"] = df3D["Vposneg"]
-        df3D = df3D_Scaled

my_static_tools.py

@@ -0,0 +1,224 @@
+from typing import Union, NamedTuple
+import io,os,re,sys,math,time,uuid,ctypes,pickle,random,shutil,string,urllib,decimal,datetime,itertools,traceback,collections,statistics
+import numpy as np, pandas as pd
+import plotly.express as px
+import huggingface_hub
+
+import sklearn #, statsmodels
+from sklearn import svm, neighbors, naive_bayes, neural_network, tree, ensemble, linear_model, discriminant_analysis, gaussian_process, manifold, cluster
+#from statsmodels.tsa import seasonal
+
+os.makedirs(".temp", exist_ok=True) # for temporary local files
+
+""" remove decoration and popup menu button at top """
+STYLE_CORRECTION = " ".join([
+    "<style>",
+    "header[data-testid='stHeader'] { display:none }",
+    "div[data-testid='stSidebarHeader'] { display:none }",
+    "div[data-testid='stAppViewBlockContainer'] { padding:1em }",
+    "div[data-testid='collapsedControl'] { background-color:#EEE }",
+    "a[href='https://streamlit.io/cloud'] { display:none }"
+    "</style>"
+])
+
+###
+def pandas_info(df: pd.DataFrame) -> Union[pd.DataFrame,str]:
+    buffer = io.StringIO()
+    df.info(buf=buffer)
+    str_info = buffer.getvalue()
+    try:
+        lines = str_info.splitlines()
+        df = (pd.DataFrame([x.split() for x in lines[5:-2]], columns=lines[3].split()).drop('Count',axis=1).rename(columns={'Non-Null':'Non-Null Count'}))
+        return df
+    except Exception as ex:
+        print(ex)
+        return str_info
+
+
+def pandas_random_dataframe(n_cols:int = 15, n_rows:int = 100) -> pd.DataFrame:
+    """ create random dataframe - случайные числа, для отладки например """
+    df = pd.DataFrame(np.random.randn(n_rows, n_cols), columns=(f"col {i}" for i in range(n_cols)))
+    return df
+
+
+########################################################################################################################################
+
+
+class HfRepo(NamedTuple):
+    repo_id: str
+    repo_type: str
+    token: str
+
+class HF_tools:
+    """ Huggingface tools """
+
+    def list_models_spaces(token: str, author = 'f64k'):
+        """ list models and spaces """
+        api = huggingface_hub.HfApi(token=token)
+        #spaces = api.list_spaces(author=author)
+        models = api.list_models(author=author)
+        datasets = api.list_datasets(author=author)
+        lstResult = list(datasets) + list(models)
+        lstResult = [ {"id": i.id, "type": type(i).__name__, "private": i.private, "tags": i.tags} for i in lstResult]
+        return lstResult
+
+    def save_dataframe_to_hf(repo: HfRepo, dfToSave: pd.DataFrame, new_filename: str, remote_subdir: str) -> Union[huggingface_hub.CommitInfo, Exception]:
+        """ save dataframe to hf repo """
+        try:
+            local_filename = os.path.join(".temp", new_filename)
+            #df.to_csv('compressed_data.zip', index=False, compression={'method': 'zip', 'archive_name': 'data.csv'})
+            dfToSave.to_csv(local_filename, index=False, sep=";", encoding="utf-8") # , compression="zip"
+            apiHF = huggingface_hub.HfApi(token=repo.token)
+            path_in_repo = os.path.basename(local_filename)
+            if remote_subdir:
+                path_in_repo = f"{remote_subdir}/{path_in_repo}"
+            commit_info = apiHF.upload_file(path_or_fileobj=local_filename, path_in_repo=path_in_repo, repo_id=repo.repo_id, repo_type=repo.repo_type)
+            return commit_info
+        except Exception as exSave:
+            return exSave
+
+    def load_dataframes_from_hf(repo: HfRepo, lstCsvFiles: list[str] = []) -> {str, pd.DataFrame}:
+        """ load dataframes from hf """
+        #https://huggingface.co/datasets/f64k/gaziev/blob/main/TestData3_2204_noAB_gaziev.zip
+        dict_res = {}
+        for fl_name in lstCsvFiles:
+            try: file_loaded = huggingface_hub.hf_hub_download(filename=fl_name, repo_id=repo.repo_id, repo_type=repo.repo_type, token=repo.token)
+            except: file_loaded = ""
+            if os.path.exists(file_loaded):
+                compress = "zip" if file_loaded.lower().endswith("zip") else None
+                df_loaded = pd.read_csv(file_loaded, sep=";", encoding = "utf-8", compression=compress)
+                dict_res[fl_name] = df_loaded # df_Vproc = df_process_v_column(df_loaded)
+        return dict_res
+
+    def list_files_hf(repo: HfRepo) -> list[str]:
+        """ List CSV and ZIP files in HF repo - список CSV и ZIP файлов (c уровнем вложенности) в репозитории """
+        ### https://huggingface.co/docs/huggingface_hub/en/guides/hf_file_system
+        fs = huggingface_hub.HfFileSystem(token=repo.token, use_listings_cache=False) # , skip_instance_cache=True
+        path_hf = f"{repo.repo_type}s/{repo.repo_id}/"
+        #lst = fs.ls(path_hf, detail=False)
+        lstGlob = fs.glob(path_hf + "**") # map(os.path.basename, lstGlob)
+        lstNames = [fname.replace(path_hf, "") for fname in lstGlob if fname.lower().endswith(".csv") or fname.lower().endswith(".zip")]
+        #print(f"ПРОЧИТАН�� В list_files_hf() : {lstNames=}")
+        return lstNames
+
+
+########################################################################################################################################
+
+
+RANDOM_STATE=11
+
+class XYZV_tools:
+    """ XYZV tools - для данных в специальном формате """
+
+    def df_process_v_column(df: pd.DataFrame) -> pd.DataFrame:
+        """ обработка столбца V для дальнейшего удобства + столб T типа время """
+        df = df.reset_index() #
+        df.rename(columns = {"index": "T"}, inplace=True)
+        df["Vis"] = df.V.map(lambda v: 0 if str(v)=="nan" else 1).astype(int)
+        df["Vfloat"] = df.V.map(lambda v: 0 if str(v)=="nan" else str(v).replace(',', '.')).astype(float)
+        df["Vsign"] = df.Vfloat.map(lambda v: -1 if v<0 else 1 if v>0 else 0).astype(int)
+        df["Vposneg"] = df.Vfloat.map(lambda v: "n" if v<0 else "p" if v>0 else "o").astype(str)
+        return df
+
+    @staticmethod
+    def CreateDictClassifiers_BestForXYZ() :
+        dictFastTree = {
+            #"RandomForestClassifier": ensemble.RandomForestClassifier(random_state=RANDOM_STATE), # совсем плохие показатели
+            #"ExtraTreeClassifier": tree.ExtraTreeClassifier(random_state=RANDOM_STATE), #
+            "DecisionTreeClassifier": tree.DecisionTreeClassifier(random_state=RANDOM_STATE), # лучший по последним баллам
+        }
+        #return {**dictFast}
+        #return {**dict_Test_MLPClassifier}
+        #return {**dictFast, **dictLongTrain}
+        return {**dictFastTree}
+
+    # lstRepoZipFiles = ["TrainData_1504_AB_gaziev.zip","TestData_1504_AB_gaziev.zip","TestData3_2204_noAB_gaziev.zip"]
+    ### returns (classifier_object, df_train_with_predict, time_elapsed)
+    def GetClassifier(lstDfOriginal, nHystorySteps) :
+        #lstDfOriginal = [df_9125_Train, df_12010_Test, df_9051_Test3]
+        nShift = nHystorySteps
+        nCurrShift = nHystorySteps
+        classifierName = "DecisionTreeClassifier"
+        colsVectorInp = ["X","Y","Z"]
+        fieldY = "Vis" #
+        lstDataFrames = XYZV_tools.MakeHystoryColumns(lstDfOriginal, nShift)
+        df_train = pd.concat(lstDataFrames)
+        lstColsShift = [f"{c}-{i}" for i in range(1, nCurrShift+1) for c in colsVectorInp] # для nCurrShift=0 lstColsShift=[]
+        colsVectorInpAll = colsVectorInp + lstColsShift
+        y_train = df_train[fieldY]
+        x_train_vect = df_train[colsVectorInpAll]
+        dictClassifiers = XYZV_tools.CreateDictClassifiers_BestForXYZ()
+        classifierObject = dictClassifiers[classifierName]
+        start2 = time.time()
+        classifierObject.fit(x_train_vect, y_train) # процесс обучения
+        time_elapsed = time.time() - start2
+        y_pred = classifierObject.predict(x_train_vect.values)  # .values[:,::-1] поля XYZ и истории в обратном порядке
+        df_train[f"predict_{fieldY}"] = y_pred
+        return (classifierObject, df_train, time_elapsed)
+
+    #
+    def MakeHystoryColumns(lstDfOriginal, nShift) :
+        lstDataframesShifted = [df.copy() for df in lstDfOriginal]
+        lstColsShift = []
+        for i in range(1, nShift+1):
+            #cols = ["X","Y","Z"]+["A","B"]
+            cols = ["X","Y","Z"]
+            #cols = ["A","B"]
+            for c in cols:
+                for dfShift in lstDataframesShifted:
+                    dfShift[f'{c}-{i}'] = dfShift[c].shift(i).fillna(0)
+                lstColsShift.append(lstDataframesShifted[0].columns[-1])
+        print(lstColsShift)
+        return lstDataframesShifted
+
+    ###
+    def plotly_xyzv_scatter_gray(df3D):
+        """ 3D plot """
+        color_discrete_map = dict(o='rgb(230,230,230)', p='rgb(90,1,1)', n='rgb(1,1,90)')
+        fig = px.scatter_3d(df3D, x='X', y='Y', z='Z', color="Vposneg", opacity=0.4, height=800, color_discrete_map=color_discrete_map)
+        fig.update_scenes(
+            xaxis={"gridcolor":"rgba(30, 0, 0, 0.2)","color":"rgb(100, 0, 0)","showbackground":False},
+            yaxis={"gridcolor":"rgba(0, 30, 0, 0.2)","color":"rgb(0, 100, 0)","showbackground":False},
+            zaxis={"gridcolor":"rgba(0, 0, 30, 0.2)","color":"rgb(0, 0, 100)","showbackground":False})
+        fig.update_traces(marker_size=3)
+        return fig
+
+
+
+########################################################################################################################################
+
+
+
+
+
+
+
+
+#import joblib
+#REPO_ID = "YOUR_REPO_ID"
+#FILENAME = "sklearn_model.joblib"
+#model = joblib.load(hf_hub_download(repo_id=REPO_ID, filename=FILENAME))
+
+
+if False:
+    if False:
+        # https://scikit-learn.org/stable/auto_examples/preprocessing/plot_all_scaling.html
+        scaler = sklearn.preprocessing.StandardScaler()
+        #scaler = sklearn.preprocessing.PowerTransformer()
+        #scaler = sklearn.preprocessing.RobustScaler()
+        #scaler = sklearn.preprocessing.MinMaxScaler() # https://scikit-learn.org/1.1/modules/generated/sklearn.preprocessing.MinMaxScaler.html#sklearn.preprocessing.MinMaxScaler
+        #scaler = sklearn.preprocessing.MinMaxScaler(feature_range=(-1,1))
+        #scaler = sklearn.preprocessing.QuantileTransformer()
+        #scaler = sklearn.preprocessing.QuantileTransformer(output_distribution="normal")
+        #scaler = sklearn.preprocessing.Normalizer() # всё на сферу кладёт - приводит к 1 длину вектора
+        scale_columns = ["X","Y","Z"]
+        scaledData = scaler.fit_transform(df3D[scale_columns])
+        if False:
+            scaler2 = sklearn.preprocessing.Normalizer()
+            scaledData = scaler2.fit_transform(scaledData)
+        df3D_Scaled = pd.DataFrame(data=scaledData, columns=scale_columns)
+        df3D_Scaled["Vposneg"] = df3D["Vposneg"]
+        df3D = df3D_Scaled
+
+
+

pages/2_TECT_IDXYZ.py

@@ -1,9 +1,12 @@
 import os, re, sys, time, math, shutil, urllib, string, random, pickle, zipfile, datetime, platform
 import streamlit as st, pandas as pd, numpy as np
-import my_static_methods as my_stm
+import my_static_tools as mst
+from my_static_tools import HF_tools as hft
+from my_static_tools import XYZV_tools as xvt
+
 
 if True:
-    st.html(my_stm.STYLE_CORRECTION)
+    st.html(mst.STYLE_CORRECTION)
     dirParams = {
         "os.getcwd": os.getcwd(),
         "cpu_count": os.cpu_count(),
@@ -23,20 +26,20 @@ if True:
     #st.sidebar.write(dirParams)
 
 
-REPO = my_stm.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
-lstRepoFiles = my_stm.list_files_hf(REPO) # список уже имеющихся в репозитории файлов
+REPO = mst.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
+lstRepoFiles = hft.list_files_hf(REPO) # список уже имеющихся в репозитории файлов
 dictTestFilesIdXyz = {f.upper().replace("ID_XYZ/",""): f.upper() for f in lstRepoFiles if f.upper().startswith("ID_XYZ/")}
 
 @st.cache_data
 def GetListOf_XYZV_ToTrainClassifier(repo):
     lstRepoZipFiles = ["TrainData_1504_AB_gaziev.zip","TestData_1504_AB_gaziev.zip","TestData3_2204_noAB_gaziev.zip"]
-    dictTrainThreeDataframes = my_stm.load_dataframes_from_hf(repo, lstRepoZipFiles)
-    lstDfOriginal = [my_stm.df_process_v_column(df) for df in dictTrainThreeDataframes.values()]
+    dictTrainThreeDataframes = hft.load_dataframes_from_hf(repo, lstRepoZipFiles)
+    lstDfOriginal = [xvt.df_process_v_column(df) for df in dictTrainThreeDataframes.values()]
     return lstDfOriginal
 
 @st.cache_data
 def GetCachedClassifier(lstDfOriginal, nHystorySteps):
-    classifier_object, df_train_with_predict, time_elapsed = my_stm.GetClassifier(lstDfOriginal, nHystorySteps)
+    classifier_object, df_train_with_predict, time_elapsed = xvt.GetClassifier(lstDfOriginal, nHystorySteps)
     #st.session_state.df_train_with_predict = df_train_with_predict
     columns_xyzv = [c for c in df_train_with_predict.columns if "Vis" in c] + [c for c in df_train_with_predict.columns if c[0] in "XYZ"]
     st.session_state.df_train_with_predict = df_train_with_predict[columns_xyzv]
@@ -67,7 +70,7 @@ def DescriptionMarkdown() -> str:
     """
 
 def save_dataframe_nodialog_idxyz(new_filename, dfToSave):
-    commit_info = my_stm.save_dataframe_to_hf(REPO, dfToSave, new_filename, "ID_XYZ")
+    commit_info = hft.save_dataframe_to_hf(REPO, dfToSave, new_filename, "ID_XYZ")
     st.toast(commit_info, icon='🆕')
     ReRun()
 
@@ -124,13 +127,13 @@ with col1.popover("🆕 добавить новый файл", use_container_wid
                 st.error(f"Столбцы не ID;X;Y;Z ! Наблюдаем столбцы : {colnames}")
 
 # список уже имеющихся в репозитории файлов. повторное чтение
-lstRepoFiles = my_stm.list_files_hf(REPO)
+lstRepoFiles = hft.list_files_hf(REPO)
 dictTestFilesIdXyz = {f.upper().replace("ID_XYZ/",""): f.upper() for f in lstRepoFiles if f.upper().startswith("ID_XYZ/")}
 selectedFile = col1.radio("📰 загруженные тестовые пакеты", dictTestFilesIdXyz.keys(), index=None)
 
 # выбран файл для предсказания
 if selectedFile is not None:
-    dict_ONE_IDXYZ = my_stm.load_dataframes_from_hf(REPO, [dictTestFilesIdXyz[selectedFile]])
+    dict_ONE_IDXYZ = hft.load_dataframes_from_hf(REPO, [dictTestFilesIdXyz[selectedFile]])
     if len(dict_ONE_IDXYZ) > 0:
         df_idxyz = list(dict_ONE_IDXYZ.values())[0]
         dfShow = df_idxyz

pages/4_Загрузка Просмотр CSV.py

@@ -1,9 +1,13 @@
 import os, csv, json
 import streamlit as st, pandas as pd, numpy as np
 import huggingface_hub
-import my_static_methods as my_stm
-st.html(my_stm.STYLE_CORRECTION)
-REPO = my_stm.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
+
+import my_static_tools as mst
+from my_static_tools import HF_tools as hft
+from my_static_tools import XYZV_tools as xvt
+
+st.html(mst.STYLE_CORRECTION)
+REPO = mst.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
 
 st.sidebar.markdown("# 💾 Загрузка")
 st.sidebar.markdown("# 🧊 Просмотр")
@@ -21,7 +25,7 @@ def save_dataframe_dialog_start(new_filename, dfToSave):
     cnt_msg = st.container()
     d_cols = st.columns(2)
     if d_cols[1].button("Сохранить"):
-        commit_info = my_stm.save_dataframe_to_hf(REPO, dfToSave, new_filename, "XYZ_CSV")
+        commit_info = hft.save_dataframe_to_hf(REPO, dfToSave, new_filename, "XYZ_CSV")
         cnt_msg.write(commit_info)
         #st_rerun()
     if d_cols[0].button("Закрыть"):
@@ -68,23 +72,23 @@ with tab1:
             if col1.button(f"можно сохранить на сервер файл '{fileXYZ}'"): # .to_excel(fileXlsName)
                 save_dataframe_dialog_start(fileXYZ, df)
         col2.write(df.describe())
-        dfinfo = my_stm.pandas_info(df)
+        dfinfo = mst.pandas_info(df)
         col2.write(dfinfo)
 
         #col1.write(df.aggregate(["mean","median","prod","sum","std","var"]))
 
-lstTestFiles = [f for f in my_stm.list_files_hf(REPO) if not f.upper().startswith("ID_XYZ/")]
+lstTestFiles = [f for f in hft.list_files_hf(REPO) if not f.upper().startswith("ID_XYZ/")]
 
 with tab2:
     cols = tab2.columns(2)
     #key_xyz = st.selectbox("Выберите файл данных для просмотра таблицы и точек", dictXYZV.keys())
     one_file_selected = cols[0].selectbox("Выберите файл данных для просмотра таблицы и точек", lstTestFiles)
     if one_file_selected:
-        dict_ONE_XYZV = my_stm.load_dataframes_from_hf(REPO, [one_file_selected])
+        dict_ONE_XYZV = hft.load_dataframes_from_hf(REPO, [one_file_selected])
         if len(dict_ONE_XYZV) > 0:
             df_xyz = list(dict_ONE_XYZV.values())[0] #df_xyz = dictXYZV[key_xyz]
-            df_xyz_vproc = my_stm.df_process_v_column(df_xyz)
-            fig = my_stm.plotly_xyzv_scatter_gray(df_xyz_vproc)
+            df_xyz_vproc = xvt.df_process_v_column(df_xyz)
+            fig = xvt.plotly_xyzv_scatter_gray(df_xyz_vproc)
             cols[0].plotly_chart(fig) # st.plotly_chart(fig) #fig.show()
             lstValues = sorted(set(df_xyz.notna().sum()), reverse=True)
             cols[1].info(f"Bceгo записей : {lstValues[0]};  Значений V : {lstValues[1]}")

pages/6_Chat.py

@@ -1,14 +1,18 @@
 import os, re, sys, time, math, shutil, urllib, string, random, pickle, zipfile, datetime
 import streamlit as st, pandas as pd, numpy as np
-import my_static_methods as my_stm
 from faker import Faker
 
+import my_static_tools as mst
+from my_static_tools import HF_tools as hft
+from my_static_tools import XYZV_tools as xvt
+
+
 # https://docs.gspread.org/en/latest/user-guide.html#updating-cells
 # https://docs.streamlit.io/develop/tutorials/databases/private-gsheet
 # https://github.com/streamlit/gsheets-connection/blob/main/examples/pages/Service_Account_Example.py#L167
 #from streamlit_gsheets import GSheetsConnection # st-gsheets-connection - конфликт с pandas 2.2.2
 
-st.html(my_stm.STYLE_CORRECTION)
+st.html(mst.STYLE_CORRECTION)
 st.sidebar.markdown("# Переговоры 💬")
 
 # Create a connection object.
@@ -16,8 +20,8 @@ st.sidebar.markdown("# Переговоры 💬")
 #st.write(conn)
 #st.help(conn)
 
-REPO = my_stm.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
-lstRepoFiles = my_stm.list_files_hf(REPO) # список уже имеющихся в репозитории файлов
+REPO = mst.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
+lstRepoFiles = hft.list_files_hf(REPO) # список уже имеющихся в репозитории файлов
 dictTestFilesIdXyz = {f.upper().replace("ID_XYZ/",""): f.upper() for f in lstRepoFiles if f.upper().startswith("ID_XYZ/")}
 
 """

pages/9_Таблица_результатов.py

@@ -1,11 +1,11 @@
 import streamlit as st, pandas as pd, numpy as np
-import my_static_methods as my_stm
-st.markdown(my_stm.STYLE_CORRECTION, unsafe_allow_html=True)
+import my_static_tools as mst
+st.markdown(mst.STYLE_CORRECTION, unsafe_allow_html=True)
 
 st.sidebar.markdown("###  просто таблица случайных чисел - пока заглушка ❄️")
 
 pop = st.popover("Open popover")
-pop.dataframe(my_stm.df_random_dataframe())
+pop.dataframe(mst.pandas_random_dataframe())
 
 expand = st.sidebar.expander("My label", icon=":material/info:")
 expand.write("Inside the expander.")

pages/old/_plotly_graph_objs.py

@@ -1,6 +1,6 @@
 import streamlit as st, pandas as pd, numpy as np
-import my_static_methods as my_stm
-st.html(my_stm.STYLE_CORRECTION)
+import my_static_tools as mst
+st.html(mst.STYLE_CORRECTION)
 
 import plotly.figure_factory as ff
 import plotly.graph_objs as go

pages/old/_просмотр_CSV.py

@@ -1,28 +1,28 @@
 import os
 import streamlit as st, pandas as pd, numpy as np
-import my_static_methods as my_stm
-st.html(my_stm.STYLE_CORRECTION)
+import my_static_tools as mst
+st.html(mst.STYLE_CORRECTION)
 
 #st.sidebar.title("⚜️")
 
 #col1.metric("Temperature", "70 °F", "1.2 °F") #col2.metric("Wind", "9 mph", "-8%") #col3.metric("Humidity", "86%", "4%")
 
 
-REPO = my_stm.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
+REPO = mst.HfRepo("f64k/gaziev", "dataset", st.secrets["HF_WRITE"])
 st.sidebar.info("🧊💽 сохраненные\n таблицы CSV")
 
-lstTestFiles = [f for f in my_stm.list_files_hf(REPO) if not f.upper().startswith("ID_XYZ/")]
+lstTestFiles = [f for f in mst.list_files_hf(REPO) if not f.upper().startswith("ID_XYZ/")]
 
 cols = st.columns(2)
 
 #key_xyz = st.selectbox("Выберите файл данных для просмотра таблицы и точек", dictXYZV.keys())
 one_file_selected = cols[0].selectbox("Выберите файл данных для просмотра таблицы и точек", lstTestFiles)
 if one_file_selected:
-    dict_ONE_XYZV = my_stm.load_dataframes_from_hf(REPO, [one_file_selected])
+    dict_ONE_XYZV = mst.load_dataframes_from_hf(REPO, [one_file_selected])
     if len(dict_ONE_XYZV) > 0:
         df_xyz = list(dict_ONE_XYZV.values())[0] #df_xyz = dictXYZV[key_xyz]
-        df_xyz_vproc = my_stm.df_process_v_column(df_xyz)
-        fig_plotly_xyzv_scatter = my_stm.plotly_xyzv_scatter_gray(df_xyz_vproc)
+        df_xyz_vproc = mst.df_process_v_column(df_xyz)
+        fig_plotly_xyzv_scatter = mst.plotly_xyzv_scatter_gray(df_xyz_vproc)
         cols[0].plotly_chart(fig_plotly_xyzv_scatter) #fig.show()
         lstValues = sorted(set(df_xyz.notna().sum()), reverse=True)
         cols[1].info(f"Bceгo записей : {lstValues[0]};  Значений V : {lstValues[1]}")