If a date is filled with empty cells or you don't want an error in calculation of a cell is zero, NULLIF is used. If cells in an alpha-numeric column is empty or null, ISNULL is used.

開発：BASICトランスパイラー。 以下の命令のリファレンスとテストが終了。 ABS ACS_ACOS ARGS ARRAY ARRAY_ADD ARRAY_AVG ARRAY_DIV ARRAY_FFT ARRAY_FFT_VIEW ARRAY_IFFT ARRAY_MAX ARRAY_MIN ARRAY_MOD ARRAY_MUL ARRAY_SUB ARRAY_SUM ASIN_ASIN ATAN2 ATN_ATAN BIN BIT BREAK BYTE CBRT CEIL CHR_CHR$ CLS CONTINUE COS DATA DATATEXT DATE DEC DEG DIM DIRS DRAW DRAW_LINE DRAW_PEN DRAW_PENSET END EOF EXP FILES FILL FILTER FILTER_BLUR FILTER_EMBOSS FILTER_HALFTONE FILTER_LUMINANCE FILTER_MOSAIC FILTER_SHARP FILTER_THRESHOLD FILTER_TONE FIX FLOOR FOR-NEXT FORMAT FRAC GOSUB GOSUBX GOTO GROUP$ HEIGHT HEX IF...THEN INPUT INT ISINF ISNULL KEY LABEL LEFT_LEFT$ LEN LET LOADIMAGE LOG LOG10 MID_MID$ OPTION_BASE OPTION_BINARY_ARRAY_TRIM OPTION_DELIMITER OPTION_ENDIAN OPTION_INPUT_PROMPT OPTION_PI OPTION_PRINT_DELIMITER OPTION_PRINT_NULL OPTION_READBIN_MAXSIZE OPTION_READBIN_MAXSIZE_ERROR OPTION_READDATA_MISSING PARAM PGET PI POW PRINT PRINT$ PRINTBIN PSET RAD READ READBIN READDATA READEXCELDATA READHTML READHTMLTEXT READLINE READTEXT REM REPEAT-UNTIL RESTORE RETURN RETURNX RIGHT_RIGHT$ RND ROUND SAVEIMAGE SAVEPDF SAVESVG SAVEWORD SCREEN SGN_SIGN SIN SIZE SPC SQR SRAND_RANDOMIZE STR SWAP TAN TIMER TRUNC TYPE VAL VALS WHILE-WEND WIDTH

NULLIF is not the same as ISNULL, they are just similar

The new "big" thing I tried in this session was keyed findings. Everything that gets written to an md file gets an ID before it gets written up. B-001 for a branch in the code, R-001 for a rule we need to lock in and test, T-001 for a table in the db, etc. Now every fact has an address. The agent reasons way better when its own findings are keyed, it can point to R-014 and check it against everything else instead of juggling "that ISNULL(...) rule on the PD branch". The keys are there for me too when I review, but mostly they keep the agent grounded. This was the biggest difference vs previous attempts of something of this scale, I think.

Every Monday morning, before any stakeholder sees a number, I run the same four checks. Not because I distrust my pipeline. Because I have been burned enough times to know trust is not a substitute for verification. The pattern in Python: import pandas as pd df = pd.read_csv('weekly_report.csv') # 1. Row count sanity check print(f"Rows this week: {len(df)}") # 2. Nulls on key columns print(df[['revenue','customer_id','date']].isnull().sum()) # 3. Date range check print(df['date'].min(), 'to', df['date'].max()) # 4. Revenue outlier flag outliers = df[df['revenue'] > df['revenue'].quantile(0.99)] if len(outliers) > 0: print(f"{len(outliers)} outlier rows detected") Two minutes. Has caught broken pipelines, missing date ranges, and one load error heading into a board pack. Senior analysts trusted by leadership are the ones who catch their own before anyone else does. What does your pre-send data check look like?

وجود القيم الفارغة قد يسبب أخطاء في الكود أو يؤدي إلى نتائج وتوقعات غير دقيقة للنماذج. لمعرفة حجم المشكلة في مكتبة Pandas، نستخدم دالة ⁠()isnull().sum⁠ لتحديد عدد القيم المفقودة في كل عمود والبدء في وضع استراتيجية للتعامل معها. #Python #DataScience

セールスフォースの数式項目で「ある項目が空欄の場合はAを、空欄ではない場合はBを返す」という式を作る場合、= null ではなく ISBLANK()かISNULL()を使う必要があります。二つの違いはこちら。 note.com/effinous/n/n136b3f4… #Salesforce

Day 5 of #60DaysOfLearning2026 DataFrame Merging & Joins Merged movies dataset with taglines using pd.merge() Learned different joins: Left, Right, Inner & Outer Handled missing values using .isnull().sum() @lftechnology #LSPPDay5 #60DaysOfLearning2026 #LearningWithLeapfrog

開発：BASICトランスパイラー。今夜の段階で説明かけたのは以下のもの。現在CodexがDEC()の仕様変更に対応してるところ。配達時間なので、ここまで。 ABS,ARGS,CEIL,CONTINUE,COS,END,FLOOR,FOR-NEXT,GOSUB,GOTO,INPUT,INT,ISNULL,LABEL,LEFT,LEN,LET,OPTION_BASE,OPTION_PI,PI,PRINT$,PRINTBIN,REM,REPEAT-UNTIL,RIGHT,ROUND,SGN_SIGN,SIN,SIZE,SQR,SRAND_RANDOMIZE,STR,SWAP,TAN,TIMER,TRUNC,TYPE,VAL,VALS,WHILE-WEND

5 Pandas Series functions every beginner should know: - isnull() — Identifies missing (NaN) values in a dataset. - fillna() — Replaces missing values with a specified value. - dropna() — Removes missing values from the Series. - astype() — Converts data from one data type to another. - apply() — Applies a custom function to every element in the Series. These are some of the most frequently used Pandas functions and form the foundation of working with real-world data.

Pandas is one of the most important Python libraries in Data Science, ML and Data Engineering. It helps you: • Read CSV/Excel files • Clean messy data • Handle missing values • Filter and sort data • Analyze large datasets • Prepare data for ML models Core structures: • Series → 1D data • DataFrame → table-like 2D data Some important beginner functions in Pandas: • read_csv() → read CSV files • head() → first 5 rows • info() → dataset summary • describe() → statistics • shape → rows & columns • columns → column names • isnull() → check missing values • dropna() / fillna() → handle null values • loc[] / iloc[] → select rows & columns • sort_values() → sorting • groupby() → data analysis • merge() → combine datasets • value_counts() → frequency count • to_csv() → save files

WHERE句にISNULLを持ってきているせいで、いらん完全NULLレコードが取得されてしまっているのでは……？ いやでも昨日まではちゃんと動いてたんだよな～～～～～ 何が悪い

機械学習を始める前にやること、知ってる？ モデルを作る前にまずこれ👇 ① データの形を確認（shape） ② 欠損値を確認（isnull） ③ 分布を確認（describe/hist） モデルより前処理が8割。 ここをサボると後で必ず詰む🔥 #機械学習 #Python #駆け出しエンジニアと繋がりたい

Day 7 Today, I learned: • What to do with missing data ( keep it, remove it, or fill it) • How to use .isnull() and .notnull() • Using groupby() to categorize data • That groupby() is often used with aggregation functions like .mean() #LearnInPublic

There is a reason why ISNULL is in SQL. NULL values doesn't always mean something is wrong or the data needs cleaning. Experience will eventually teach you that.

【5/13 本日の学習】 所感：新しいメソッドを学ぶことができたよかった。 メソッドを利用したその裏でなにが起きているかも合わせて理解することが重要実感 ≪　学習内容　≫ ① DataFrame → 表全体のデータを扱う ② Series → DataFrame の1列データを扱う ③ bool型 → True / False を扱う型 ④ isnull() df["score"].isnull() → 欠損値（NaN）を True / False で判定するメソッド NaN → True 値あり → False ⑤ isnull().sum() df["score"].isnull().sum() → 欠損値の合計数を確認できる True = 1 False = 0 として計算されるため、 True の数 = 欠損値の数になる ⑥ fillna() df["score"] = df["score"].fillna(0) → 欠損値（NaN）を指定した値で補完するメソッド 例： NaN → 0 に置き換えている。

📚 DS独学 Day 2 ・欠損値の確認（isnull・fillna） ・データ型の変換（astype） ・円グラフの作成（matplotlib） ・糖尿病データセットで実践 欠損値の確認と対処、データ分析っぽくて嬉しい。 学習時間：1.5時間 / 累計：2.5時間 #Python #データサイエンス #100DaysOfCode

IsNull(മുഖ്യമന്ത്രി ) = True

Poor teacher just needed an ISNULL handler, and it could be whatever number she wanted