Short-term solar data stream of 23-24 cycle

These datasets contain records of daily solar data as well as data collected from magnetic classes. The datasets were assembled with data from ftp://ftp.swpc.noaa.gov/pub/warehouse/. The date the data was assembled is 2017-01-15 (yyyy-mm-dd). The original data source is provided by the Space Weather Prediction Center - SWPC, which is linked to the National Oceanic and Atmospheric Administration - NOAA from US Department of Commerce. The data collected refer to the period between january 01, 1997 to january 15, 2017. Features included: <strong>radio_flux_10.7cm</strong>: the solar radio flux at 10.7 cm (2800 MHz) is an indicator of solar activity. It is also called the F10.7 index and is one of the longest running records of solar activity. Radio emissions originate high in the chromosphere and low in the corona of the solar atmosphere. <strong>sesc_sunspot_number</strong>: it refers to the number of sunspots computed on a given day. Also called Wolf's number of sunspots, it is given by R = k(10g + s), where k is a scalable factor indicating the combined effects of observation conditions, g is the number of active regions and s the number of sunspots in all these groups. <strong>sunspot_area</strong>: it refers to the sum of the corrected area of all observed sunspots. It is measured in units of millionths of the solar hemisphere. <strong>goes15_xray_bkgd_flux</strong>: it corresponds to the daily average background X-ray flux that is measured by the SWPC primary GOES satellite. To calculate this value, sensors register 24 X-ray measures for a given day, one for each hour. Then, the SWPC creates 3 groups of periods of 8 hours. For these groups, the SWPC registers the lowest values of flux, creating 3 minimal values, one for each group. Then, they calculate the average between the minimum values of the first and the third group. After the average calculation, they must compare this value to the minimal value of the second group. The minimum value from the last comparison gives the result of the X-ray background flux. <strong>mwl_alpha</strong>: binary attribute indicating the existence of apha magnetic class in any observed spot. <strong>mwl_beta</strong>: binary attribute indicating the existence of beta magnetic class in any observed spot. <strong>mwl_gamma</strong>: binary attribute indicating the existence of gamma magnetic class in any observed spot. <strong>mwl_beta_gamma</strong>: binary attribute indicating the existence of beta-gamma magnetic class in any observed spot. <strong>mwl_beta_delta</strong>: binary attribute indicating the existence of beta-delta magnetic class in any observed spot. <strong>mwl_beta_gamma_delta</strong>: binary attribute indicating the existence of beta-gamma-delta magnetic class in any observed spot. <strong>mwl_gamma_delta</strong>: binary attribute indicating the existence of gamma-delta magnetic class in any observed spot. <strong>mwl_delta</strong>: binary attribute indicating the existence of delta magnetic class in any observed spot. We performed missing data imputation using k-NN over all features. The k-NN used the Gower's distance as its distance coefficient. In addition, we also performed z-score standardization in all features. We designed the data into a sliding time window stream. In other words, we designed the data stream regarding four days before a<em> </em><em>t1 </em>instant (i.e. <em>t5, t4, t3,</em> and <em>t2</em>). Hence, new features were created considering the evolution of data along five days: <em>radio_flux_10.7cm_[t5, t4, t3, t2, t1];</em> <em>sesc_sunspot_number_[t5, t4, t3, t2, t1];</em> <em>sunspot_area_[t5, t4, t3, t2, t1];</em> <em>goes15_xray_bkgd_flux_[t5, t4, t3, t2, t1];</em> <em>mwl_alpha_[t5, t4, t3, t2, t1];</em> <em>mwl_beta_[t5, t4, t3, t2, t1];</em> <em>mwl_gamma_[t5, t4, t3, t2, t1];</em> <em>mwl_beta_gamma_[t5, t4, t3, t2, t1];</em> <em>mwl_beta_delta_[t5, t4, t3, t2, t1];</em> <em>mwl_beta_gamma_delta_[t5, t4, t3, t2, t1];</em> <em>mwl_gamma-delta_[t5, t4, t3, t2, t1];</em> <em>mwl_delta_[t5, t4, t3, t2, t1].</em> We designed our target variable as being the occurrence of at least one flare phenomenon of M or X class in the next 24, 24-48, and 48-72 hours ahead of the <em>t1</em> instant: <em>flare_t1d</em>: occurrence of at least one flare of class M or X in the next 24 hours ahead of the <em>t1 </em>instant; <em>flare_t2d</em>: occurrence of at least one flare of class M or X 24-48 hours ahead of the <em>t1 </em>instant; <em>flare_t3d</em>: occurrence of at least one flare of class M or X 48-72 hours ahead of the <em>t1 </em>instant.本数据集涵盖每日太阳观测记录与磁分类采集数据，整合自ftp://ftp.swpc.noaa.gov/pub/warehouse/ 处获取的原始数据。数据整合完成日期为2017年1月15日（格式为yyyy-mm-dd）。本数据集的原始数据源来自空间天气预报中心（Space Weather Prediction Center, SWPC），该机构隶属于美国商务部下辖的国家海洋和大气管理局（National Oceanic and Atmospheric Administration, NOAA）。所采集的数据覆盖1997年1月1日至2017年1月15日的时间段。 数据集包含以下特征： - 10.7厘米射电通量（radio_flux_10.7cm）：指10.7厘米（2800 MHz）波段的太阳射电通量，是表征太阳活动的核心指标之一，也被称为F10.7指数，是持续时间最长的太阳活动观测记录之一。该射电辐射源自太阳大气的色球层高层与日冕低层。 - 太阳黑子数（sesc_sunspot_number）：指单日统计的太阳黑子总数，也被称为沃尔夫太阳黑子数，计算公式为$R = k(10g + s)$，其中$k$为表征观测条件综合影响的缩放因子，$g$为活动区数量，$s$为所有活动群中的太阳黑子个数。 - 太阳黑子总面积（sunspot_area）：指所有观测到的太阳黑子的校正面积之和，单位为太阳半球的百万分之一。 - GOES-15 X射线背景通量（goes15_xray_bkgd_flux）：指空间天气预报中心（SWPC）的主GOES卫星测得的每日平均X射线背景通量。计算该值时，传感器会在单日每小时记录一次X射线通量，共获取24组数据。随后SWPC将单日时间划分为3个8小时时段，分别记录每个时段内的最低通量值，得到3个极小值。接着计算第一与第三时段极小值的平均值，并将该平均值与第二时段的极小值进行比较，最终取二者中的较小值作为X射线背景通量结果。 - 磁分类α（mwl_alpha）：二元属性，用于表征任意观测黑子中是否存在α磁分类。 - 磁分类β（mwl_beta）：二元属性，用于表征任意观测黑子中是否存在β磁分类。 - 磁分类γ（mwl_gamma）：二元属性，用于表征任意观测黑子中是否存在γ磁分类。 - 磁分类β-γ（mwl_beta_gamma）：二元属性，用于表征任意观测黑子中是否存在β-γ复合磁分类。 - 磁分类β-δ（mwl_beta_delta）：二元属性，用于表征任意观测黑子中是否存在β-δ复合磁分类。 - 磁分类β-γ-δ（mwl_beta_gamma_delta）：二元属性，用于表征任意观测黑子中是否存在β-γ-δ复合磁分类。 - 磁分类γ-δ（mwl_gamma_delta）：二元属性，用于表征任意观测黑子中是否存在γ-δ复合磁分类。 - 磁分类δ（mwl_delta）：二元属性，用于表征任意观测黑子中是否存在δ磁分类。 我们对所有特征采用k近邻（k-NN）算法进行缺失值填充，该算法以高沃距离（Gower's distance）作为距离度量系数。此外，我们还对所有特征进行了z-score标准化处理。 我们将数据构建为滑动时间窗口数据流。具体而言，我们以时刻$t_1$为基准，选取其前4天的数据（即$t_5$、$t_4$、$t_3$与$t_2$）构建数据流。由此，我们基于5天的时间序列演化生成新特征，包括：radio_flux_10.7cm_[t5, t4, t3, t2, t1]、sesc_sunspot_number_[t5, t4, t3, t2, t1]、sunspot_area_[t5, t4, t3, t2, t1]、goes15_xray_bkgd_flux_[t5, t4, t3, t2, t1]、mwl_alpha_[t5, t4, t3, t2, t1]、mwl_beta_[t5, t4, t3, t2, t1]、mwl_gamma_[t5, t4, t3, t2, t1]、mwl_beta_gamma_[t5, t4, t3, t2, t1]、mwl_beta_delta_[t5, t4, t3, t2, t1]、mwl_beta_gamma_delta_[t5, t4, t3, t2, t1]、mwl_gamma-delta_[t5, t4, t3, t2, t1]以及mwl_delta_[t5, t4, t3, t2, t1]。 我们将目标变量设置为：在时刻$t_1$之后的24小时、24至48小时以及48至72小时内，是否至少发生一次M级或X级耀斑事件，具体包括： - flare_t1d：时刻$t_1$之后的24小时内至少发生一次M级或X级耀斑事件； - flare_t2d：时刻$t_1$之后的24至48小时内至少发生一次M级或X级耀斑事件； - flare_t3d：时刻$t_1$之后的48至72小时内至少发生一次M级或X级耀斑事件。