final commit

2025_03_21/1_interpolation.py

@@ -0,0 +1,20 @@
+import pandas as pd
+
+# Load the full CSV (with all columns)
+df = pd.read_csv("in/df_suppressed.csv", encoding='UTF-8', on_bad_lines='skip', delimiter=';')  # or 'cp1252' if needed
+
+# Replace "X" and "Y" with your actual column names if needed
+x_col = 'Lambert_X'
+y_col = 'Lambert_Y'
+
+# Ensure X and Y are numeric, coerce invalid to NaN
+df[x_col] = pd.to_numeric(df[x_col], errors='coerce')
+df[y_col] = pd.to_numeric(df[y_col], errors='coerce')
+
+# Interpolate only the X and Y columns (other columns untouched)
+df[[x_col, y_col]] = df[[x_col, y_col]].interpolate(method='linear', limit_direction='both')
+
+# Save to a new CSV
+df.to_csv("out/interpolated_trace.csv", index=False, sep=';')
+
+print("✅ Interpolation done on X/Y. Other columns preserved.")

2025_03_21/2_to_shp.py

@@ -0,0 +1,24 @@
+import pandas as pd
+import geopandas as gpd
+from shapely.geometry import Point
+
+# 1. Load the interpolated CSV
+df = pd.read_csv("out/interpolated_trace.csv", encoding='UTF-8', on_bad_lines='skip', delimiter=';')  # or 'cp1252' if needed
+
+# 2. Define your coordinate column names (replace if needed)
+x_col = 'Lambert_X'
+y_col = 'Lambert_Y'
+
+# 3. Create geometry from X and Y
+geometry = [Point(xy) for xy in zip(df[x_col], df[y_col])]
+
+# 4. Create GeoDataFrame
+gdf = gpd.GeoDataFrame(df, geometry=geometry)
+
+# 5. Set CRS (Lambert 93 = EPSG:2154, common in France)
+gdf.set_crs(epsg=2154, inplace=True)
+
+# 6. Export to shapefile
+gdf.to_file("out/interpolated_trace.shp")
+
+print("✅ Shapefile created: interpolated_trace.shp")

2025_03_21/csv_to_shp.py

@@ -0,0 +1,23 @@
+import pandas as pd
+import geopandas as gpd
+from shapely.geometry import Point
+import sys
+
+def csv_to_shp(in_file: str, out_file: str, x_col: str = 'Lambert_X', y_col: str = 'Lambert_Y'):
+    # 1. Load the interpolated CSV
+    df = pd.read_csv(in_file, encoding='UTF-8', on_bad_lines='skip', delimiter=';')  # or 'cp1252' if needed
+
+    # 3. Create geometry from X and Y
+    geometry = [Point(xy) for xy in zip(df[x_col], df[y_col])]
+
+    # 4. Create GeoDataFrame
+    gdf = gpd.GeoDataFrame(df, geometry=geometry)
+
+    # 5. Set CRS (Lambert 93 = EPSG:2154, common in France)
+    gdf.set_crs(epsg=2154, inplace=True)
+
+    # 6. Export to shapefile
+    gdf.to_file(out_file)
+    
+csv_to_shp(sys.argv[1], sys.argv[2], sys.argv[3], sys.argv[4])
+

2025_03_21/full.py

@@ -0,0 +1,93 @@
+import pandas as pd
+import geopandas as gpd
+from shapely.geometry import Point
+import numpy as np
+import pyproj
+import math
+import sys
+
+# Function to convert WGS84 to Lambert 93 for the entire DataFrame
+def convert_to_lambert93(row, transf):
+    if(math.isnan(row[x_col]) or math.isnan(row[y_col])):
+        x, y = transf.transform(row[x_inter_col], row[y_inter_col])
+        row[x_inter_col] = x
+        row[y_inter_col] = y
+    return row
+
+def point_position_on_line(pt):
+    return line.project(pt)
+
+# Load CSV
+df = pd.read_csv(sys.argv[1], encoding='UTF-8', on_bad_lines='skip', delimiter=';')
+
+x_col = sys.argv[4]
+y_col = sys.argv[5]
+
+x_inter_col = sys.argv[6]
+y_inter_col = sys.argv[7]
+
+df[x_col] = pd.to_numeric(df[x_col], errors='coerce')
+df[y_col] = pd.to_numeric(df[y_col], errors='coerce')
+
+# Load shapefile (must contain LineString)
+line_gdf = gpd.read_file(sys.argv[2])
+line = line_gdf.union_all()  # merge if multiple lines
+
+# Create geometry
+df['geometry'] = df.apply(
+    lambda row: Point(row[x_col], row[y_col]) if not pd.isna(row[x_col]) and not pd.isna(row[y_col]) else None,
+    axis=1
+)
+
+# Get position
+df['distance_on_line'] = df['geometry'].apply(lambda g: line.project(g) if g else np.nan)
+
+# Interpolate
+df['distance_on_line'] = df['distance_on_line'].interpolate(method='linear', limit_direction='both')
+
+# Create interpolated points
+df['geometry'] = df['distance_on_line'].apply(lambda d: line.interpolate(d))
+
+# Start with original values
+df[x_inter_col] = df[x_col]
+df[y_inter_col] = df[y_col]
+
+# Update only missing ones
+df.loc[df[x_col].isna(), x_inter_col] = df.loc[df[x_col].isna(), 'geometry'].apply(lambda g: g.x)
+df.loc[df[y_col].isna(), y_inter_col] = df.loc[df[y_col].isna(), 'geometry'].apply(lambda g: g.y)
+
+
+# Load CSV and shape
+df = pd.read_csv("in/raw.csv", delimiter=';', encoding='utf-8')
+df[x_col] = pd.to_numeric(df[x_col], errors='coerce')
+df[y_col] = pd.to_numeric(df[y_col], errors='coerce')
+
+# Load the shapefile (must be a LineString)
+line = gpd.read_file("in/célé.shp").unary_union  # Merge multiple lines if needed
+
+# Assign geometry to known points
+df['geometry'] = df.apply(
+    lambda r: Point(r[x_col], r[y_col]) if pd.notna(r[x_col]) else None,
+    axis=1
+)
+
+# Compute the position of known points along the line
+df['distance_on_line'] = df['geometry'].apply(lambda g: line.project(g) if g else np.nan)
+
+# Interpolate missing distances (ensures every point has a valid position)
+df['distance_on_line'] = df['distance_on_line'].interpolate(method='linear', limit_direction='both')
+
+# Interpolate new coordinates from the reference shape
+df['geometry'] = df['distance_on_line'].apply(lambda d: line.interpolate(d) if pd.notna(d) else None)
+df[x_inter_col] = df['geometry'].apply(lambda g: g.x if g else np.nan)
+df[y_inter_col] = df['geometry'].apply(lambda g: g.y if g else np.nan)
+
+# Apply the conversion to the entire DataFrame row-wise
+wgs84 = pyproj.CRS("EPSG:4326")
+lambert93 = pyproj.CRS("EPSG:2154")
+transformer = pyproj.Transformer.from_crs(wgs84, lambert93, always_xy=True)
+
+df = df.apply(lambda row : convert_to_lambert93(row, transformer), axis=1)
+
+# Save fixed CSV
+df.to_csv(sys.argv[3], sep=';', index=False)

2025_03_21/in/célé.cpg

@@ -0,0 +1 @@
+UTF-8

2025_03_21/in/célé.prj

@@ -0,0 +1 @@
+GEOGCS["GCS_WGS_1984",DATUM["D_WGS_1984",SPHEROID["WGS_1984",6378137.0,298.257223563]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]]

2025_03_21/in/célé.qmd

@@ -0,0 +1,43 @@
+<!DOCTYPE qgis PUBLIC 'http://mrcc.com/qgis.dtd' 'SYSTEM'>
+<qgis version="3.22.16-Bia?owie?a">
+  <identifier></identifier>
+  <parentidentifier></parentidentifier>
+  <language></language>
+  <type>dataset</type>
+  <title></title>
+  <abstract></abstract>
+  <contact>
+    <name></name>
+    <organization></organization>
+    <position></position>
+    <voice></voice>
+    <fax></fax>
+    <email></email>
+    <role></role>
+  </contact>
+  <links/>
+  <fees></fees>
+  <encoding></encoding>
+  <crs>
+    <spatialrefsys>
+      <wkt></wkt>
+      <proj4></proj4>
+      <srsid>0</srsid>
+      <srid>0</srid>
+      <authid></authid>
+      <description></description>
+      <projectionacronym></projectionacronym>
+      <ellipsoidacronym></ellipsoidacronym>
+      <geographicflag>false</geographicflag>
+    </spatialrefsys>
+  </crs>
+  <extent>
+    <spatial crs="" minx="0" miny="0" minz="0" dimensions="2" maxz="0" maxx="0" maxy="0"/>
+    <temporal>
+      <period>
+        <start></start>
+        <end></end>
+      </period>
+    </temporal>
+  </extent>
+</qgis>

2025_03_21/log.txt

@@ -0,0 +1 @@
+Interpolation done on X/Y. Other columns preserved.

2025_03_21/out/interpolated_with_trace.cpg

@@ -0,0 +1 @@
+UTF-8

2025_03_21/out/interpolated_with_trace.prj

@@ -0,0 +1 @@
+PROJCS["RGF_1993_Lambert_93",GEOGCS["GCS_RGF_1993",DATUM["D_RGF_1993",SPHEROID["GRS_1980",6378137.0,298.257222101]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic"],PARAMETER["False_Easting",700000.0],PARAMETER["False_Northing",6600000.0],PARAMETER["Central_Meridian",3.0],PARAMETER["Standard_Parallel_1",49.0],PARAMETER["Standard_Parallel_2",44.0],PARAMETER["Latitude_Of_Origin",46.5],UNIT["Meter",1.0]]

2025_03_21/out/raw.cpg

@@ -0,0 +1 @@
+UTF-8

2025_03_21/out/raw.prj

@@ -0,0 +1 @@
+PROJCS["RGF_1993_Lambert_93",GEOGCS["GCS_RGF_1993",DATUM["D_RGF_1993",SPHEROID["GRS_1980",6378137.0,298.257222101]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Lambert_Conformal_Conic"],PARAMETER["False_Easting",700000.0],PARAMETER["False_Northing",6600000.0],PARAMETER["Central_Meridian",3.0],PARAMETER["Standard_Parallel_1",49.0],PARAMETER["Standard_Parallel_2",44.0],PARAMETER["Latitude_Of_Origin",46.5],UNIT["Meter",1.0]]

2025_03_21/temp.py

@@ -0,0 +1,96 @@
+# -*- coding: utf-8 -*-
+
+import pyproj
+import os
+import pandas as pd
+
+directory_path = 'Y:\MISSIONS\Eau\8 - Projet recherche Célé\Lucie\Science\Canoo\continuum Ce guillaume 2020\dossier_brut_2020'
+
+dataframes = []
+
+for filename in os.listdir(directory_path):
+    if filename.endswith('.csv'):
+        file_path = os.path.join(directory_path, filename)
+        df = pd.read_csv(file_path, delimiter=',', skiprows=18, parse_dates=[0])
+        dataframes.append(df)
+
+merged_dataframe = pd.concat(dataframes, ignore_index=True)
+merged_dataframe.sort_values(by=['Date Heure'], inplace=True)
+merged_dataframe.reset_index(drop=True, inplace=True)
+
+
+
+columns_to_suppress_indices = [4, 5, 7, 9, 10, 11, 12, 21, 22]
+
+if max(columns_to_suppress_indices) >= len(merged_dataframe.columns):
+    print("Invalid column index found.")
+else:
+    df_suppressed = merged_dataframe.drop(merged_dataframe.columns[columns_to_suppress_indices], axis=1)
+   
+#supprimer les données manquantes
+#df_suppressed.dropna(subset=['Latitude (°)', 'Longitude (°)'], inplace=True)  
+
+# Define the coordinate systems
+wgs84 = pyproj.CRS("EPSG:4326")
+lambert93 = pyproj.CRS("EPSG:2154")
+
+# Create the Transformer to perform the conversion
+transformer = pyproj.Transformer.from_crs(wgs84, lambert93, always_xy=True)
+
+# Function to convert WGS84 to Lambert 93 for the entire DataFrame
+def convert_to_lambert93(row):
+    x, y = transformer.transform(row["Longitude (°)"], row["Latitude (°)"])
+    row["Lambert_X"] = x
+    row["Lambert_Y"] = y
+    return row
+
+# Apply the conversion to the entire DataFrame row-wise
+df_suppressed = df_suppressed.apply(convert_to_lambert93, axis=1)
+
+print(df_suppressed)
+
+
+
+#alors il faut ensuit calculer l'hypothénuse entre chacun de mes points : peut être résolu de façon booléenne
+def distance_entre_points(x1, y1, x2, y2):
+    distance = ((x2 - x1)**2 + (y2 - y1)**2) ** 0.5
+    return distance
+
+# Créer une nouvelle colonne 'Distance' dans le DataFrame
+df_suppressed['Distance'] = 0.0
+
+# Créer une nouvelle colonne 'Distance Cumulative' pour la somme cumulative des distances
+df_suppressed['Distance Cumulative'] = 0.0
+
+# Variable pour garder la somme cumulative des distances
+cumulative_distance = 0.0
+
+    
+for i in range(len(df_suppressed) - 3):
+    x1, y1 = df_suppressed.loc[i, 'Lambert_X'], df_suppressed.loc[i, 'Lambert_Y']
+    x2, y2 = df_suppressed.loc[i + 1, 'Lambert_X'], df_suppressed.loc[i + 1, 'Lambert_Y']
+    distance = distance_entre_points(x1, y1, x2, y2)
+    df_suppressed.loc[i + 1, 'Distance'] = distance
+    cumulative_distance += distance
+    df_suppressed.loc[i + 1, 'Distance Cumulative'] = cumulative_distance
+
+# Afficher le DataFrame avec les colonnes 'Distance' et 'Distance Cumulative' mises à jour
+print(df_suppressed)
+
+df_suppressed.to_csv('Y:\MISSIONS\Eau\8 - Projet recherche Célé\Lucie\continuum\continuum lucie + fabs + steph\df_suppressedbrut_bon_dernier.csv', index=False)
+#faire les graohs des paramètres en fonction de la distance  = variations des paramètres
+
+#Essayer tous les paramètres 
+#comparer avec les données de 2020. 
+#corriger la pression de la baro.
+
+#Donc changement de stratégie. Je vais essayer de retrouver les données GPS manquante par interpolation de celle-ci.
+#Pour cela on va faire un travail en sept étapes:
+#    1. Sur Qgis exporter un fichier csv. du cours d'eau sous forme de points avec les coordonnées GPS en Lambert 93. 
+#    2. J'importe ce nouveau csv dans Python et je calcule la distance entre chaque points sur toute ma rivière. 
+ #   3. A partir de là je calcul la distance cumulative entre chacun de mes points sur tous mon cours d'eau.
+ #   4. Ensuite je regarde à la mano là où j'ai des trous. Je calcul sur Qgis la distance manquante de point GPS 
+  #  5. Distance manquante / nombres de mesures = distance entre chaque points de mesure (exemple 2m)
+   # 6. Je calcul la distance cumulative au niveau de mes trous 
+    #7. Je vais chercher dans le tableau 1 le coordonnée GPS correspondant à la distance cumulative similaire. 
+    #8. Le tour est joué.