Skip to content

Commit

Permalink
fix some linter warning and maluse of double quotes
Browse files Browse the repository at this point in the history
  • Loading branch information
@chraibi
chraibi committed Sep 29, 2024
1 parent 044411d commit c61becc
Showing 1 changed file with 41 additions and 105 deletions.
146 changes: 41 additions & 105 deletions src/tabs/LargeView_analysis.py
View file
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
""" This script computes the density and velocity fields from the trajectories of pedestrians. """
"""This script computes the density and velocity fields from the trajectories of pedestrians."""

import logging
import pickle
Expand Down Expand Up @@ -117,9 +117,7 @@ def gaussian_kernel_scalar(r: float, r_c: float, xi: float) -> float:


@njit
def get_r(
i_line: int, j_column: int, r_cg: float, x_min: float, y_min: float
) -> Tuple[float, float]:
def get_r(i_line: int, j_column: int, r_cg: float, x_min: float, y_min: float) -> Tuple[float, float]:
"""
Calculate the coordinates (x, y) of a point in a grid based on its indices and grid parameters.
Expand Down Expand Up @@ -155,9 +153,7 @@ def get_cell(r: tuple, x_min: float, y_min, r_cg: float) -> Tuple[int, int]:
return (i_line, j_column)


def butter_lowpass_filter(
pd_series: pd.Series, delta_t: float, order: int, cutoff: float
) -> np.ndarray:
def butter_lowpass_filter(pd_series: pd.Series, delta_t: float, order: int, cutoff: float) -> np.ndarray:
"""
Applies a Butterworth lowpass filter to a pandas Series.
Expand All @@ -173,9 +169,7 @@ def butter_lowpass_filter(
nyquist_freq = 0.5 / delta_t # Nyquist Frequency
normal_cutoff = cutoff / nyquist_freq # Normalized cutoff frequency
# Get the filter coefficients
b, a = butter(
order, normal_cutoff, btype="low", analog=False
) # Generate the filter coefficients
b, a = butter(order, normal_cutoff, btype="low", analog=False) # Generate the filter coefficients
y = filtfilt(b, a, pd_series, padlen=int(1 / delta_t) + 1) # Apply the filter.
# This function performs forward and backward filtering to eliminate phase distortion,
# ensuring that the output signal has no phase shift relative to the input.
Expand Down Expand Up @@ -210,10 +204,7 @@ def create_save_folder(pa: Parameters) -> Path:
Path: The path of the created save folder.
"""
save_folder = (
pa.FOLDER_SAVE
/ f"{pa.SELECTED_NAME}_start={pa.START_TIME:.1f}_dur={pa.DURATION:.1f}_dt={pa.DELTA_T:.1f}_xi={pa.XI:.1f}_rcg={pa.R_CG:.1f}"
)
save_folder = pa.FOLDER_SAVE / f"{pa.SELECTED_NAME}_start={pa.START_TIME:.1f}_dur={pa.DURATION:.1f}_dt={pa.DELTA_T:.1f}_xi={pa.XI:.1f}_rcg={pa.R_CG:.1f}"
save_folder.mkdir(parents=True, exist_ok=True)
return save_folder

Expand Down Expand Up @@ -293,10 +284,7 @@ def process_trajectories(all_datas: pd.DataFrame, pa: Parameters) -> Dict[str, p
traj_data.loc[:, "y_m"] = (1.0 - alpha) * Y_bw + alpha * traj_data["y_m"].values

# Filter the data to the desired time interval
traj_data = traj_data[
(traj_data["t_s"] >= pa.START_TIME)
& (traj_data["t_s"] < pa.START_TIME + pa.DURATION + 2.0 * pa.DT)
]
traj_data = traj_data[(traj_data["t_s"] >= pa.START_TIME) & (traj_data["t_s"] < pa.START_TIME + pa.DURATION + 2.0 * pa.DT)]

# If the trajectory is empty, skip it
if traj_data.empty:
Expand Down Expand Up @@ -398,12 +386,7 @@ def initialize_dict(nb_cg_x: int, nb_cg_y: int) -> Dict[str, np.ndarray]:
# List of field names
field_names = ["X", "Y", "rho", "vxs", "vys", "vxs2", "vys2", "var_vs"]

# Initialize arrays using a dictionary comprehension
density_velocity_fields = {
name: np.zeros((nb_cg_x, nb_cg_y), dtype="d") for name in field_names
}

return density_velocity_fields
return {name: np.zeros((nb_cg_x, nb_cg_y), dtype="d") for name in field_names}


def compute_fields(
Expand All @@ -425,10 +408,7 @@ def compute_fields(
nb_traj = len(all_trajs)
# Iterate over all trajectories
for i_traj, traj in tqdm(enumerate(all_trajs.values()), desc="Processing grid"):
traj = traj.loc[
(traj["t_s"] >= pa.START_TIME)
& (traj["t_s"] < pa.START_TIME + pa.DURATION + 2.0 * pa.DT)
]
traj = traj.loc[(traj["t_s"] >= pa.START_TIME) & (traj["t_s"] < pa.START_TIME + pa.DURATION + 2.0 * pa.DT)]
if traj.shape[0] == 0:
continue
# Iterate over all rows in the trajectory (ie all time steps)
Expand All @@ -438,18 +418,14 @@ def compute_fields(
# Iterate over the cells in the grid
for i in range(i_rel - pa.DELTA, i_rel + pa.DELTA + 1):
for j in range(j_rel - pa.DELTA, j_rel + pa.DELTA + 1):
if (
i < 0 or i >= pa.NB_CG_X or j < 0 or j >= pa.NB_CG_Y
): # Check if pedestrian is outside the grid
if i < 0 or i >= pa.NB_CG_X or j < 0 or j >= pa.NB_CG_Y: # Check if pedestrian is outside the grid
continue

# Compute the position of the grid cell
pos_grid_cell = get_r(i, j, pa.R_CG, pa.X_MIN, pa.Y_MIN)

# Compute the Gaussian kernel
phi_r = gaussian_kernel_scalar(
calculate_distance(pos_grid_cell, pos_ped), pa.R_C, pa.XI
)
phi_r = gaussian_kernel_scalar(calculate_distance(pos_grid_cell, pos_ped), pa.R_C, pa.XI)

# Update the fields
df_observables["X"][i, j] = pos_grid_cell[0]
Expand Down Expand Up @@ -502,9 +478,7 @@ def normalize_density(density: np.ndarray, pa: Parameters) -> np.ndarray:
return density


def update_figure(
df_observables: Dict[str, np.ndarray], pa: Parameters, plot_density: bool, zsmooth=False
) -> go.Figure:
def update_figure(df_observables: Dict[str, np.ndarray], pa: Parameters, plot_density: bool, zsmooth=False) -> go.Figure:
"""
Update the figure with the given observables and parameters.
Expand All @@ -518,21 +492,14 @@ def update_figure(
"""
# Create the X and Y axes
X_axis = np.linspace(
pa.X_MIN, pa.X_MIN + len(df_observables["rho"][0]) * pa.R_CG, len(df_observables["rho"][0])
)
Y_axis = np.linspace(
pa.Y_MIN, pa.Y_MIN + len(df_observables["rho"]) * pa.R_CG, len(df_observables["rho"])
)
X_axis = np.linspace(pa.X_MIN, pa.X_MIN + len(df_observables["rho"][0]) * pa.R_CG, len(df_observables["rho"][0]))
Y_axis = np.linspace(pa.Y_MIN, pa.Y_MIN + len(df_observables["rho"]) * pa.R_CG, len(df_observables["rho"]))

# Create the heatmap
if plot_density:
hovertext = np.array(
[
[
f"x: {X_axis[i]:.2f} m<br>y: {Y_axis[j]:.2f} m<br>Density: {df_observables["rho"][j,i]:.2f} ped/m²"
for i in range(df_observables["rho"].shape[1])
]
[f"x: {X_axis[i]:.2f} m<br>y: {Y_axis[j]:.2f} m<br>Density: {df_observables['rho'][j,i]:.2f} ped/m²" for i in range(df_observables["rho"].shape[1])]
for j in range(df_observables["rho"].shape[0])
]
)
Expand All @@ -546,18 +513,15 @@ def update_figure(
zmin=pa.COLORBAR_MIN, # Set the minimum value of the colorbar
zmax=pa.COLORBAR_MAX, # Set the maximum value of the colorbar
zsmooth=zsmooth, # Apply smoothing with 'best' interpolation
colorbar=dict(title="Density [ped/m²]", titleside="right"),
colorbar={"title": "Density [ped/m²]", "titleside": "right"},
hoverinfo="text", # Set hoverinfo to display just the hovertext
hovertext=hovertext, # Include hovertext
)

else:
hovertext = np.array(
[
[
f"x: {X_axis[i]:.2f} m<br>y: {Y_axis[j]:.2f} m<br>Var_v: {(df_observables["var_vs"][j,i]):.2f} m²/s²"
for i in range(df_observables["var_vs"].shape[1])
]
[f"x: {X_axis[i]:.2f} m<br>y: {Y_axis[j]:.2f} m<br>Var_v: {(df_observables['var_vs'][j,i]):.2f} m²/s²" for i in range(df_observables["var_vs"].shape[1])]
for j in range(df_observables["var_vs"].shape[0])
]
)
Expand All @@ -571,7 +535,7 @@ def update_figure(
zmin=pa.COLORBAR_MIN, # Set the minimum value of the colorbar
zmax=pa.COLORBAR_MAX_V, # Set the maximum value of the colorbar
zsmooth=zsmooth, # Apply smoothing with 'best' interpolation
colorbar=dict(title="Var_v [m²/s²]", titleside="right"),
colorbar={"title": "Var_v [m²/s²]", "titleside": "right"},
hoverinfo="text", # Set hoverinfo to display just the hovertext
hovertext=hovertext, # Include hovertext
)
Expand All @@ -585,7 +549,7 @@ def update_figure(
scale=pa.QUIVER_SCALE,
hoverinfo="skip",
arrow_scale=0.2,
line=dict(width=1, color="black"),
line={"width": 1, "color": "black"},
)

# Combine both plots in a single figure
Expand Down Expand Up @@ -617,38 +581,26 @@ def update_figure(
)

# Add text annotation for the arrow
fig.add_annotation(
x=9.5, y=3.8, text="<b>1 m/s</b>", showarrow=False, font=dict(color="#009999", size=25)
)
fig.add_annotation(x=9.5, y=3.8, text="<b>1 m/s</b>", showarrow=False, font={"color": "#009999", "size": 25})

# Correct layout update with proper domain
fig.update_layout(
font=dict(
# family="Courier New, monospace",
size=20,
),
title=dict(
text=(
"Velocity field for density heatmap"
if plot_density
else "Velocity field for variance velocity heatmap"
),
font_size=20,
),
xaxis=dict(
title=dict(text="x [m]", font_size=20),
scaleanchor="y",
scaleratio=1,
range=[0.0, 12.0],
tickfont_size=20,
),
yaxis=dict(
title=dict(text="y [m]", font_size=20),
scaleanchor="x",
scaleratio=1,
range=[2.2, 23],
tickfont_size=20,
),
font={"size": 20},
title={"text": ("Velocity field for density heatmap" if plot_density else "Velocity field for variance velocity heatmap"), "font_size": 20},
xaxis={
"title": {"text": "x [m]", "font_size": 20},
"scaleanchor": "y",
"scaleratio": 1,
"range": [0.0, 12.0],
"tickfont_size": 20,
},
yaxis={
"title": {"text": "y [m]", "font_size": 20},
"scaleanchor": "x",
"scaleratio": 1,
"range": [2.2, 23],
"tickfont_size": 20,
},
width=650,
height=900,
)
Expand Down Expand Up @@ -702,16 +654,10 @@ def main(selected_file: str):
QUIVER_SCALE=3.0,
)
pa.R_CG = slider_value_R_CG
pa.DELTA = (
int(ceil(pa.R_C / pa.R_CG)) + 1
) # Number of cells to consider around the cell containing the point
pa.DELTA = int(ceil(pa.R_C / pa.R_CG)) + 1 # Number of cells to consider around the cell containing the point
folder_save = create_save_folder(pa)

if (
not Path(folder_save / "traj_data.pkl").exists()
or not Path(folder_save / "parameters.pkl").exists()
or not Path(folder_save / "dictionnary_observables.pkl").exists()
):
if not Path(folder_save / "traj_data.pkl").exists() or not Path(folder_save / "parameters.pkl").exists() or not Path(folder_save / "dictionnary_observables.pkl").exists():
########## PROGRESS BAR ##########
title_text = st.text("Progress Bar")
my_progress_bar = st.progress(0)
Expand Down Expand Up @@ -750,26 +696,16 @@ def main(selected_file: str):
with col1:
fig = update_figure(dict_observables, params, True, st.session_state["zsmooth"])
st.plotly_chart(fig)
figname = (
pa.SELECTED_NAME
+ f"_velocity_field_for_density_heatmap_zsmooth{st.session_state["zsmooth"]}.pdf"
)
figname = pa.SELECTED_NAME + f"_velocity_field_for_density_heatmap_zsmooth{st.session_state['zsmooth']}.pdf"
img_bytes = fig.to_image(format="pdf")
st.sidebar.download_button(
label="Download Density Heatmap", data=img_bytes, file_name=figname
)
st.sidebar.download_button(label="Download Density Heatmap", data=img_bytes, file_name=figname)
plt.close()
with col2:
fig = update_figure(dict_observables, params, False, st.session_state["zsmooth"])
st.plotly_chart(fig)
figname = (
pa.SELECTED_NAME
+ f"_velocity_field_for_density_heatmap_zsmooth{st.session_state["zsmooth"]}.pdf"
)
figname = pa.SELECTED_NAME + f"_velocity_field_for_density_heatmap_zsmooth{st.session_state['zsmooth']}.pdf"
img_bytes = fig.to_image(format="pdf")
st.sidebar.download_button(
label="Download Variance Heatmap", data=img_bytes, file_name=figname
)
st.sidebar.download_button(label="Download Variance Heatmap", data=img_bytes, file_name=figname)
plt.close()


Expand Down

0 comments on commit c61becc

Please sign in to comment.