superset用戶使用流程手冊

....Superset1地圖，平行坐標(biāo)，熱力圖，箱線圖，樹狀圖，熱力圖，水平圖等圖，官網(wǎng)上是不行以操作多個(gè)表的，不過我們可以操作視圖，也就是說在數(shù)據(jù)庫建好視superset以去配置緩存，來加速儀盤表的查詢，不必要沒次都去查詢數(shù)據(jù)庫。3、我們可以查看進(jìn)展查詢表的sql，也可以把查詢導(dǎo)出為json，csv文件。它sqlsql。一、系統(tǒng)根本使用在使用本系統(tǒng)前，應(yīng)領(lǐng)先進(jìn)展賬戶的建立操作。登陸輸入開通的賬號、密碼點(diǎn)擊登錄即可。建賬號先登陸admin用戶，點(diǎn)擊界面欄目“安全“→“用戶列表“→，點(diǎn)擊按鈕在登陸時(shí)可以使用先登陸admin用戶，點(diǎn)擊界面欄目“安全“→“用戶列表“→，點(diǎn)擊按鈕在填寫完信息后點(diǎn)擊保存〔注：紅色*的選項(xiàng)未必填項(xiàng)〕即可使用建賬號登陸。在創(chuàng)立賬號時(shí)，依據(jù)使用賬號功能需求選擇不同的角色屬性，每個(gè)賬號的用戶名不能一樣。二、創(chuàng)立數(shù)據(jù)源supersetMySQLsuperset進(jìn)入superset的Web界面，點(diǎn)擊sources下拉選項(xiàng)的Databases，如以以下圖：進(jìn)入數(shù)據(jù)庫界面，點(diǎn)擊“+”按鈕進(jìn)入數(shù)據(jù)庫連接界面，填寫正確字段后保存，操作如下：查看已經(jīng)連接好的數(shù)據(jù)庫添加數(shù)據(jù)表Sources->Tables點(diǎn)擊加號〔+〕增數(shù)據(jù)表Database選擇之前創(chuàng)立好的數(shù)據(jù)源，TableNameDateTimestamp數(shù)據(jù)表查看、編輯Sources->Tables可對表構(gòu)造、數(shù)據(jù)類型、是否可進(jìn)展group、filter、count、sum、min、max操作等進(jìn)展編輯數(shù)據(jù)探究分析與可視化呈現(xiàn)Table定義好維度字段和指標(biāo)之后，即可針對該表進(jìn)展數(shù)據(jù)探究分析與可視化呈現(xiàn)，在ListTable并可以選擇相應(yīng)的圖表進(jìn)展可視化呈現(xiàn)。可視化圖表類型選擇選擇可視化圖表類型。Superset自帶的圖表類型如上，包括柱狀圖、餅圖、時(shí)間序列線圖、積存圖、圖表、熱詞圖等。數(shù)據(jù)時(shí)間范圍選擇選擇：時(shí)間所在數(shù)據(jù)列。時(shí)間粒度，時(shí)間跨度維度呈現(xiàn)選擇、圖表可視化選擇Groupby：xMetrics：y軸呈現(xiàn)的數(shù)據(jù)指標(biāo)〔包括指標(biāo)的sum、avg等〕SortBy：排序依據(jù)可視化選擇配色圖例是否積存／分布……坐標(biāo)軸編輯選擇X、Y編輯X、YY自定義查詢／過濾自定義SQL語句where、havingin／notin查詢可視化保存sliceslicedashboarddashboardDashboard編輯每個(gè)Slice對應(yīng)的模塊，可以自由拖拽位置和大小，并保存整個(gè)Dashboard的布局。多表關(guān)聯(lián)查詢在數(shù)據(jù)表編輯界面，可通過DatabaseExpression，運(yùn)用SQL語句實(shí)現(xiàn)多表關(guān)聯(lián)查詢。Superset經(jīng)過上邊的步驟就連接上了數(shù)據(jù)庫，下邊就可以進(jìn)展數(shù)據(jù)的可視化操作了。首先點(diǎn)擊SQLSQLSQL點(diǎn)擊Visualize按鈕進(jìn)入數(shù)據(jù)可視化編輯窗口：supersetDistribution–BarChart〔分布-條形圖〕案例需求：統(tǒng)計(jì)每個(gè)state的總?cè)藬?shù)，總女生人數(shù)，總男生人數(shù)。SELECTSELECTstateASstate,sum(num)ASsumnum,sum(sum_girls)ASsumsum_girls,sum(sum_boys)ASsumsum_boysFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:21:31”ANDds<=”2023-07-0518:21:31”GROUPBYstateORDERBYsumnumDESCLIMIT50000進(jìn)入可視化界面，可按需求顯示圖形：在圖形的右上方有對圖形的保存等一些操作：圖表的保存：查看保存的圖表：TableView–表視圖SELECTnameASSELECTnameASname,genderASgender,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:24:31”ANDds<=”2023-07-0518:24:31”GROUPBYname,genderORDERBYsumnumDESCLIMIT50000PivotTable–數(shù)據(jù)透視表數(shù)據(jù)透視表〔PivotTable〕是一種交互式的表，可以進(jìn)展某些計(jì)算，如求和與計(jì)數(shù)等。所進(jìn)展的計(jì)算與數(shù)據(jù)跟數(shù)據(jù)透視表中的排列有關(guān)。SELECTSELECTgenderASgender,stateASstate,nameASname,sum(num)ASsumnumFROMFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:25:21”ANDds<=”2023-07-0518:25:21”GROUPBYgender,state,nameORDERBYsumnumDESCLIMIT50000TimeSeries–LineChart–時(shí)序線圖SELECTstateASSELECTstateASstate,dsAStimestamp,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryINNERJOIN(SELECTstateASstate,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:26:35”ANDds<=”2023-07-0518:26:35”GROUPBYstateORDERBYsumnumDESCLIMIT50)ASanon_1ONstate=state WHEREds>=”1917-07-0518:26:35”ANDds<=”2023-07-0518:26:35”GROUPBYstate,dsORDERBYsumnumDESCLIMIT50000SELECTSELECTstateASstate,dsAStimestamp,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryINNERJOIN(SELECTstateASstate,sum(num)ASsumnumTimeSeries–Stacked–時(shí)序面積圖示隨時(shí)間而變化的產(chǎn)生的數(shù)據(jù)可以繪制在面積圖中以強(qiáng)調(diào)總數(shù)據(jù)量。案例需求：依據(jù)每個(gè)state每年的總?cè)藬?shù)的時(shí)序圖-疊圖。SQL：FROMFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:27:06”ANDds<=”2023-07-0518:27:06”GROUPBYstateORDERBYsumnumDESCLIMIT50)ASanon_1ONstate=state WHEREds>=”1917-07-0518:27:06”ANDds<=”2023-07-0518:27:06”GROUPBYstate,dsORDERBYsumnumDESCLIMIT50000TimeSeries–BarChart–時(shí)序柱形圖SELECTstateASSELECTstateASstate,dsAStimestamp,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryINNERJOIN(SELECTstateASstate,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryWHEREstateNOTIN(”other”)ANDds>=”1917-07-0518:28:57”ANDds<=”2023-07-0518:28:57”GROUPBYstateORDERBYsumnumDESCLIMIT50)ASanon_1ONstate=state WHEREds>=”1917-07-0518:28:57”ANDds<=”2023-07-0518:28:57”ANDstateNOTIN(”other”)GROUPBYstate,dsORDERBYsumnumDESCLIMIT50000Distribution–NVD3PieChart–餅圖SELECTstateSELECTstateASstate,sum(num)ASsumnumFROM(select*frombirth_names)ASexpr_qryWHEREds>=”1917-07-0518:29:51”ANDds<=”2023-07-0518:29:51”ANDstateNOTIN(”other”)GROUPBYstateORDERBYsumnumDESCLIMIT50000BubbleChart–氣泡圖SQLSELECTcountry_nameAScountry_name,regionASregion,SUM(`SP_POP_TOTL`)AS`sumSP_POP_TOTL`,SUM(`SP_RUR_TOTL_ZS`)AS`sumSP_RUR_TOTL_ZS`,SUM(`SP_DYN_LE00_IN`)AS`sumSP_DYN_LE00_IN`FROMwb_health_populationWHEREyear>=STR_TO_DATE(”2023-01-0100:00:00”,”%%Y-%%m-%%d%%H:%%i:%%s”)ANDyear<=STR_TO_DATE(”2023-01-0200:00:00”,”%%Y-%%m-%%d%%H:%%i:%%s”)ANDcountry_codeNOTIN(”TCA”,”MNP”,”DMA”,”MHL”,”MCO”,”SXM”,”CYM”,”TUV”,”IMY”,”KNA”,”ASM”,”ADO”,”AMA”,”PLW”)GROUPGROUPBYcountry_name,regionORDERBY`sumSP_POP_TOTL`DESCLIMIT50000MarKup–標(biāo)記圖4.10WordClould–文字云案例需求：顯示全部的name，且看到使用這個(gè)名字的人數(shù)比重。S