tools/cord/main.py

from matplotlib import pyplot as plt
import pandas as pd
import numpy as np
import io
import streamlit as st
# import scienceplots

# plt.style.use(['nature', 'no-latex',"cjk-sc-font"])
plt.rcParams['font.family'] = 'sans-serif'
plt.rcParams['font.sans-serif'] = ['SimHei']  # 设置中文字体
plt.rcParams['axes.unicode_minus'] = False  # 解决负号显示问题

def cubic_bezier_with_zero_derivatives(p0, p1, t_array, influence_factor):
    """
    创建三次贝塞尔曲线，确保起点和终点的导数为0
    
    参数:
    p0: 起点 [x0, y0]
    p1: 终点 [x1, y1]
    t_array: 参数数组 (0到1)
    influence_factor: 影响因子，控制控制点的位置
    
    返回:
    x_array, y_array: 贝塞尔曲线上的点
    """
    x0, y0 = p0
    x1, y1 = p1
    
    # 计算控制点，确保起点和终点导数为0
    # 控制点位置基于影响因子和两点间距离
    dx = x1 - x0
    
    # 第一个控制点：在起点右侧，y坐标与起点相同（确保起点导数为0）
    p1_control = [x0 + dx * influence_factor[0], y0]
    
    # 第二个控制点：在终点左侧，y坐标与终点相同（确保终点导数为0）
    p2_control = [x1 - dx * influence_factor[1], y1]
    
    # 计算贝塞尔曲线
    x_bezier = ((1-t_array)**3 * x0 + 
                3*(1-t_array)**2 * t_array * p1_control[0] + 
                3*(1-t_array) * t_array**2 * p2_control[0] + 
                t_array**3 * x1)
    
    y_bezier = ((1-t_array)**3 * y0 + 
                3*(1-t_array)**2 * t_array * p1_control[1] + 
                3*(1-t_array) * t_array**2 * p2_control[1] + 
                t_array**3 * y1)
    
    return x_bezier, y_bezier

# @st.cache_resource
def plot_reaction_coordinate(changed=None, _lines=None):
    """
    绘制反应坐标图
    """
    # if changed is not None:
    #     if _lines is None:
    #         raise ValueError("Lines must be provided when changing a slider.")
    #     lines = _lines
    #     i = changed//2
    #     x,y = lines[i].get_data()
    #     p1=x[0],y[0]
    #     p2=x[-1],y[-1]
    #     print(INFLU_FACTORS[i*2+1:i*2+3])
    #     line = cubic_bezier_with_zero_derivatives(p1,p2, np.linspace(0, 1, 300), INFLU_FACTORS[i*2+1:i*2+3])
    #     # lines[i].set_data([],[])
    #     lines[i].set_data(line[0], line[1])
    #     return
    
    
    
    lines = []
    fig,ax1 = plt.subplots(figsize=(9, 6))

    last=(-1,-1)
    
    maxy = data["Energy"].max()
    miny = data["Energy"].min()
    varyy = maxy - miny
    
    for i in range(data.shape[0]):
        line:pd.Series = data.loc[i]
        if last == (-1,-1):
            last = (1, line["Energy"])
            if not pd.isna(line["Name"]):
                ax1.annotate(str(line["Name"]), (1, line["Energy"]+varyy*0.03), ha='center')
        else:
            p1 = last[0]+2,line["Energy"]
            x,y = cubic_bezier_with_zero_derivatives(last,p1, np.linspace(0, 1, 300), INFLU_FACTORS[(i*2-2):i*2])
            l = ax1.plot(x, y, "-", color="black")[0]
            lines.append(l)
            if not pd.isna(line["Name"]):
                p = p1[0],p1[1]+varyy*0.03
                ax1.annotate(str(line["Name"]), p, ha='center')
            last = p1
        
        

    ax1.set_xlabel("Reaction Coordinate")
    ax1.xaxis.set_ticks([])
    ax1.set_ylabel("Energy (Hartree)")
    ax1.set_ylim(miny-varyy*0.1, maxy+varyy*0.1)
    return fig,lines

# 创建图形和坐标轴

def callback_gen(x):
    def callback():
        global INFLU_FACTORS
        INFLU_FACTORS[x-1] = st.session_state.get(f'slider_{x}', 0.5)
        plot_reaction_coordinate(changed=x, _lines=lines)

    return callback

def on_save():
    global out_file
    # for slider in slides:
    # slider.ax.set_visible(False)
    plt.draw()
    plt.tight_layout()
    out_file = io.BytesIO()
    fig.savefig(out_file, format=st.session_state.get("export_format", ".tiff")[1:], dpi=300, bbox_inches='tight')
    out_file.seek(0)
    return out_file.getvalue()

@st.cache_resource
def load_data(file):
    # 读取数据文件
    if file is not None:
        try:
            data = pd.read_excel(file) if file.name.endswith((".xlsx", ".xls")) else pd.read_csv(file)
        except Exception as e:
            st.error(f"Error reading file: {e}")
            exit()
    else:
        exit()

    num_factors = data.shape[0] * 2
    INFLU_FACTORS = [0.5] * data.shape[0] * 2  # 动态创建数组


    return data, INFLU_FACTORS


out_file = io.BytesIO()
st.set_page_config(
    page_title="反应坐标绘制",
    page_icon=":chart_with_upwards_trend:",
    layout="centered",
    initial_sidebar_state="expanded"
)
st.title("反应坐标绘制")
st.write("---")
col1,col2 = st.columns([0.7,0.3],gap="medium")

with col1:
    file = st.file_uploader("上传能量文件", type=["xlsx", "xls", "csv"])
    data, INFLU_FACTORS = load_data(file)

    fig,lines = plot_reaction_coordinate()
    stfig = st.pyplot(fig,False)
    st.selectbox("导出文件拓展名",[".tiff",".pdf",".png",".pgf"],key="export_format")
    st.download_button(
            label="Download Plot",
            data=on_save(),
            file_name="reaction_coordinate"+st.session_state.get("export_format", ".tiff"),
            # mime="image/tiff"
        )
with col2:
    st.write("调整滑块以改变反应坐标图曲线形状。")
    for i in range(data.shape[0]):
        if i!=0:
            st.slider(
                f'{data.loc[i,"Name"]} 左',
                0.0, 1.0, value=INFLU_FACTORS[i*2-1],
                key=f'slider_{i*2}',
                on_change=callback_gen(i*2)
            )
        if i!= data.shape[0] - 1:
            st.slider(
                f'{data.loc[i,"Name"]} 右',
                0.0, 1.0, value=INFLU_FACTORS[i*2],
                key=f'slider_{i*2+1}',
                on_change=callback_gen(i*2+1)
            )

st.write("---")
st.dataframe(data)