Dubhe/dubhe-visual-server/backend/backend/component/Exception/exception_read.py

# -*- coding: UTF-8 -*-
#  Copyright 2020 Zhejiang Lab. All Rights Reserved.
#
#  Licensed under the Apache License, Version 2.0 (the "License");
#  you may not use this file except in compliance with the License.
#  You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
#  Unless required by applicable law or agreed to in writing, software
#  distributed under the License is distributed on an "AS IS" BASIS,
#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
#  See the License for the specific language governing permissions and
#  limitations under the License.
#  =============================================================
import numpy as np
import math
from .histogram import Histogram


class Exception_read:
    def __init__(self, data=None, step=None):
        self.data = data
        self.step = step

    def get_histogram(self):
        for _data in self.data:
            if _data['step'] == self.step:
                # [min,max,buckets]
                res = [Histogram(_data['value']).get_data()]
                return res

    def get_flatten_shape(self, k):
        n = math.floor(pow(k / 2, 0.5))
        if n == 0:
            return 1, k

        while k % n != 0:
            n = n - 1
        m = k // n
        return n, m

    def flatten(self, data):
        shapes = data.shape
        l = len(shapes)
        if l == 1 or l == 2:
            return data
        else:
            if l == 3:
                n, m = self.get_flatten_shape(shapes[0])  # [B,W,H]
            else:
                data = data.transpose(0, 3, 1, 2)  # [B,C,W,H]
                shapes = data.shape
                n, m = self.get_flatten_shape(shapes[0] * shapes[1])
            data = data.reshape(n, m, shapes[-2], shapes[-1])  # [n,m,w,h]
            data = data.transpose(0, 2, 1, 3)  # [n,w,m,h]
            data = data.reshape(n * shapes[-2], m * shapes[-1])  # [n*w,m*h]
            return data

    def box_exception(self, data, limit=None):
        # 异常点
        data = self.flatten(data)
        up = max(limit)
        down = min(limit)
        if (up == np.nanmax(data)) and (down == np.nanmin(data)):
            return [], []

        ep = np.argwhere((data > up) | (data < down) | (np.isnan(data)))
        exception = data[ep[:, 0]].tolist() if len(data.shape) == 1 \
            else data[ep[:, 0], ep[:, 1]].tolist()
        return exception, ep.tolist()

    def box_line(self, data):
        x = data.ravel()
        # 四分位数
        p = (0, 5, 10, 25, 50, 75, 90, 95, 100)
        p_val = np.nanpercentile(x, p, interpolation='midpoint', overwrite_input=True).tolist()
        q1 = p_val[3]
        q2 = p_val[4]
        q3 = p_val[5]
        iqr = q3 - q1
        up = min(q3 + 1.5 * iqr, float(np.nanmax(x)))
        down = max(q1 - 1.5 * iqr, float(np.nanmin(x)))
        box_info = [up, q3, q2, q1, down]
        del p_val[3:6]
        return box_info, p_val

    def get_meta_data(self):
        if self.data:
            _data = self.data
        else:
            return None
        steps = []
        box_infos = []
        for i, items in enumerate(_data):
            steps.append(items['step'])
            box_info = self.box_line(items['value'])
            box_infos.append(box_info)
        res = {
            'step': steps,
            'box': box_infos
        }
        return res

    def get_data(self):
        for _data in self.data:
            if _data['step'] == self.step:
                data = _data['value']
                _shape = data.shape
                _min = data.min().tolist()
                _max = data.max().tolist()
                _mean = data.mean().tolist()
                # 变换数据数组大小到n*m
                flatten_data = self.flatten(data)
                res = [_shape, _min, _max, _mean, flatten_data.tolist()]
                return res

    def get_outlier(self, limit):
        for _data in self.data:
            if _data['step'] == self.step:
                data = _data['value']
                res = self.box_exception(data, limit)
                return res